Anova comparison of datasets

ANOVA comparison of all datasets

We are assessing the comparability of cows across all provided data-sets for inclusion in a compiled data-set. A data-set of all cows in all datasets will be used to draw random samplings of cows that represent pens on a dairy. These simulated pens will be used to fit the distribution of Dry Matter Intake (DMI) and other group level parameters will train a model that predict this distribution function. Before we can combine these data-sets together it is important to validate that the cows across the different data-sets are comparable metabolically, so that the assessment of DMI and the plotting of DMI distribution across the constructed pen groups is appropriate.

ANOVA compares two or more groups for a difference in group means. This is performed by comparing the variance between groups against the variance within each group. If the variance between groups is significantly larger than the variance within each group then the groups are considered significantly different.

The means of the values for milk production (kg), DMI (kg) and feed efficiency (fat corrected milk(FCM) (kg)/DMI (kg)) by data-set will be compared using ANOVA statistical analysis. Pairwise comparison of the group means, and visual assessment of the marginal values of each of these three outcomes will be used to assess the comparability of the data-sets and guide decision criteria for inclusion in a collated data-set.

Data description

Number of cows in all data-sets

Table 1.

Milk production

This plot demonstrates how milk production (kg) changes by week across the lactation for all cows in each data-set. Figure 1.

Dry matter intake

This plot demonstrates how DMI (kg) changes by week across the lactation for all cows in each data-set. Figure 2.

Feed Efficiency

This plot demonstrates how Feed Efficiency (kg fat corrected milk/kg DMI)) changes by week across the lactation for all cows in each data-set. Figure 3.

Box-plots of milk yield (kg) by week for each data-set

Figure-set 4.

Week 1

Week 2

Week 3

Week 4

Week 5

Week 6

Week 7

Week 8

Week 9

Week 10

Week 11

Week 12

Week 13

Week 14

Week 15

Week 16

Week 17

Week 18

Week 19

Week 20

Week 21

Week 22

Week 23

Week 24

Week 25

Week 26

Week 27

Week 28

Week 29

Week 30

Week 31

Week 32

Week 33

Week 34

Week 35

Week 36

Week 37

Week 38

Week 39

Week 40

Week 41

Week 42

Week 43

Week 44

Box-plots of DMI (kg) by week for each data-set

Figure-set 5.

Week 1

Week 2

Week 3

Week 4

Week 5

Week 6

Week 7

Week 8

Week 9

Week 10

Week 11

Week 12

Week 13

Week 14

Week 15

Week 16

Week 17

Week 18

Week 19

Week 20

Week 21

Week 22

Week 23

Week 24

Week 25

Week 26

Week 27

Week 28

Week 29

Week 30

Week 31

Week 32

Week 33

Week 34

Week 35

Week 36

Week 37

Week 38

Week 39

Week 40

Week 41

Week 42

Week 43

Week 44

Box-plots of feed efficiency by week for each data-set

Figure-set 6.

Week 1

Week 2

Week 3

Week 4

Week 5

Week 6

Week 7

Week 8

Week 9

Week 10

Week 11

Week 12

Week 13

Week 14

Week 15

Week 16

Week 17

Week 18

Week 19

Week 20

Week 21

Week 22

Week 23

Week 24

Week 25

Week 26

Week 27

Week 28

Week 29

Week 30

Week 31

Week 32

Week 33

Week 34

Week 35

Week 36

Week 37

Week 38

Week 39

Week 40

Week 41

Week 42

Week 43

Week 44

Hypothesis

\(H_0:\) the data-sets are equal in 1. milk production (kg), 2. DMI (kg), 3. feed efficiency
\(H_1:\) at least one mean is different between all data-sets

ANOVA methods

Assumptions

Variable type

Milk production (kg), DMI (kg) and feed efficiency (ratio) are all considered continuous variables. Groups to compare are data-set of cow origin which is a qualitative variable.

Independence

Between group observations are independent as data was collected from separate trials and sites by data-set. Within group observations are assumed independent as observations are on the level of cow with one measurement per time-point per cow and assumed no behavior interaction between cows as feed data was collected via individual feed bunks (Calan gates).

Normality

Residuals will be assessed for normal distribution. Each group sample size is \(n > 30\).

Equality of variance

Group variance differs (a significant levene test indicates unequal variance), but since data-set sample sizes are similar (Table 1) it is assumed that the ANOVA test is robust enough to accept this violation of the assumption. (Or use the Welch ANOVA which does not require this assumption).

Outliers

Outliers were assessed by box-plot descriptive analysis (Figure-set 4, 5, 6). No outliers were detected in milk production (kg) or DMI (kg).
In the feed efficiency variable an outlier of 9.18 was observed at week 35 for data-set 8.

This cow was traced back to the original database of that trial and a value of 10.9 was input as the milk fat yield (kg) at week 35. This is a biological impossibility and a data entry error so the feed efficiency value for this record at week 35 was removed from the ANOVA analysis.

Updated box-plot of week 35 feed efficiency

Min.	1st Qu.	Median	Mean	3rd Qu.	Max.	NA’s
0	1.076	1.267	1.225	1.406	2.412	266

Model equation

\[Y_{i,j} = μ + \beta_i+ϵ_{i,j}\]

\(Y_{i,j}\) is the \(j\)-th observation \((j = 1, 2, ..., n_i\)) of the \(i\)-th group (\(i = 1, 2, ..., 9\) levels).
\(μ\) is the common effect for the whole experiment.
\(\beta_i\) is the \(i\)-th treatment effect.
\(ϵ_{i,j}\) is the random error present in the \(j\)-th observation on the \(i\)-th treatment.

\(\sum \beta_i = 0, i = 1, ..., 9\) are considered fixed parameters and this is a fixed effects model.

Model fitting for milk (kg) as output variable

Week 1

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	6454.75	6	1075.79	27.65	< .001
Residuals	13148.65	338	38.90
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 338) = 27.65, p < .001; Eta2 = 0.33, 95% CI [0.26, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-2.4730757	-5.7667414	0.820590	0.2837237
3-1	5.7892787	1.8556481	9.722909	0.0003369
5-1	-0.1632628	-4.0076961	3.681170	0.9999997
6-1	-1.9877696	-6.1427492	2.167210	0.7913137
7-1	-9.0693725	-12.6657757	-5.472969	0.0000000
9-1	3.5124395	0.0156290	7.009250	0.0481612
3-2	8.2623544	4.5373514	11.987357	0.0000000
5-2	2.3098130	-1.3208709	5.940497	0.4903336
6-2	0.4853061	-3.4727305	4.443343	0.9998156
7-2	-6.5962968	-9.9632411	-3.229353	0.0000003
9-2	5.9855153	2.7251652	9.245865	0.0000021
5-3	-5.9525415	-10.1723853	-1.732698	0.0007184
6-3	-7.7770483	-12.2816315	-3.272465	0.0000106
7-3	-14.8586513	-18.8538396	-10.863463	0.0000000
9-3	-2.2768392	-6.1826167	1.628938	0.5969480
6-5	-1.8245068	-6.2514118	2.602398	0.8851151
7-5	-8.9061098	-12.8135063	-4.998713	0.0000000
9-5	3.6757023	-0.1402269	7.491632	0.0674953
7-6	-7.0816030	-11.2949076	-2.868298	0.0000204
9-6	5.5002091	1.3715890	9.628829	0.0018143
9-7	12.5818121	9.0158951	16.147729	0.0000000

NULL

Week 2

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	9955.76	6	1659.29	39.13	< .001
Residuals	14331.10	338	42.40
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 338) = 39.13, p < .001; Eta2 = 0.41, 95% CI [0.34, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.4234040	-2.015173	4.8619806	0.8829569
3-1	5.2486805	1.141982	9.3553785	0.0033377
5-1	7.3196574	3.306081	11.3332337	0.0000025
6-1	2.1950976	-2.142688	6.5328833	0.7441524
7-1	-9.0167665	-12.771400	-5.2621329	0.0000000
9-1	7.9261956	4.275536	11.5768549	0.0000000
3-2	3.8252765	-0.063615	7.7141680	0.0572462
5-2	5.8962534	2.105831	9.6866760	0.0001137
6-2	0.7716936	-3.360484	4.9038715	0.9979425
7-2	-10.4401705	-13.955250	-6.9250912	0.0000000
9-2	6.5027917	3.098997	9.9065868	0.0000007
5-3	2.0709769	-2.334527	6.4764805	0.8046694
6-3	-3.0535829	-7.756354	1.6491878	0.4645121
7-3	-14.2654470	-18.436411	-10.0944830	0.0000000
9-3	2.6775151	-1.400104	6.7551346	0.4502685
6-5	-5.1245598	-9.746235	-0.5028849	0.0189467
7-5	-16.3364239	-20.415734	-12.2571143	0.0000000
9-5	0.6065382	-3.377280	4.5903564	0.9993546
7-6	-11.2118641	-15.610541	-6.8131873	0.0000000
9-6	5.7310980	1.420832	10.0413645	0.0018688
9-7	16.9429621	13.220156	20.6657683	0.0000000

NULL

Week 3

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	11937.27	8	1492.16	32.24	< .001
Residuals	17770.30	384	46.28
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 384) = 32.24, p < .001; Eta2 = 0.40, 95% CI [0.34, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.5117112	-1.2664130	6.289835	0.4928067
3-1	4.9429649	0.4307448	9.455185	0.0198883
4-1	7.0005781	2.7348716	11.266285	0.0000171
5-1	10.2860927	5.8761897	14.695996	0.0000000
6-1	4.0151806	-0.7509464	8.781308	0.1784502
7-1	-7.5312819	-11.6566725	-3.405891	0.0000009
8-1	-1.0946566	-13.6766831	11.487370	0.9999991
9-1	10.0497551	6.0386060	14.060904	0.0000000
3-2	2.4312538	-1.8416522	6.704160	0.6987146
4-2	4.4888670	0.4771532	8.500581	0.0155998
5-2	7.7743815	3.6096679	11.939095	0.0000004
6-2	1.5034694	-3.0367466	6.043685	0.9824696
7-2	-10.0429931	-13.9051741	-6.180812	0.0000000
8-2	-3.6063678	-16.1045671	8.891832	0.9928851
9-2	7.5380440	3.7981359	11.277952	0.0000000
4-3	2.0576132	-2.6519256	6.767152	0.9109067
5-3	5.3431278	0.5025960	10.183659	0.0182519
6-3	-0.9277844	-6.0949373	4.239369	0.9997622
7-3	-12.4742468	-17.0570790	-7.891415	0.0000000
8-3	-6.0376216	-18.7769659	6.701723	0.8650396
9-3	5.1067902	0.6265200	9.587060	0.0125022
5-4	3.2855146	-1.3260870	7.897116	0.3931188
6-4	-2.9853976	-7.9387390	1.967944	0.6278776
7-4	-14.5318600	-18.8721909	-10.191529	0.0000000
8-4	-8.0952348	-20.7493651	4.558896	0.5473280
9-4	3.0491770	-1.1827188	7.281073	0.3770414
6-5	-6.2709121	-11.3489613	-1.192863	0.0043054
7-5	-17.8173746	-22.2995019	-13.335247	0.0000000
8-5	-11.3807493	-24.0842137	1.322715	0.1202560
9-5	-0.2363376	-4.6135438	4.140869	1.0000000
7-6	-11.5464625	-16.3794932	-6.713432	0.0000000
8-6	-5.1098372	-17.9413114	7.721637	0.9465239
9-6	6.0345746	1.2986843	10.770465	0.0026958
8-7	6.4366253	-6.1708968	19.044147	0.8085068
9-7	17.5810370	13.4906168	21.671457	0.0000000
9-8	11.1444118	-1.4261921	23.715016	0.1292001

NULL

Week 4

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	13213.25	8	1651.66	36.09	< .001
Residuals	18395.42	402	45.76
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 402) = 36.09, p < .001; Eta2 = 0.42, 95% CI [0.35, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.2424098	-1.5135821	5.9984017	0.6403778
3-1	5.3759946	0.8902071	9.8617822	0.0065401
4-1	7.8468478	3.6061298	12.0875658	0.0000006
5-1	10.0730838	5.6890140	14.4571535	0.0000000
6-1	4.3054981	-0.4327088	9.0437051	0.1088288
7-1	-7.8443076	-11.9455316	-3.7430836	0.0000002
8-1	-1.0733611	-6.4851743	4.3384521	0.9995031
9-1	10.3718147	6.3841629	14.3594665	0.0000000
3-2	3.1335848	-1.1142905	7.3814602	0.3441393
4-2	5.6044380	1.6162249	9.5926511	0.0005052
5-2	7.8306740	3.6903573	11.9709907	0.0000003
6-2	2.0630883	-2.4505311	6.5767077	0.8875458
7-2	-10.0867174	-13.9262738	-6.2471610	0.0000000
8-2	-3.3157709	-8.5320796	1.9005378	0.5565525
9-2	8.1294049	4.4114052	11.8474046	0.0000000
4-3	2.4708531	-2.2110972	7.1528035	0.7786910
5-3	4.6970891	-0.1150868	9.5092651	0.0619267
6-3	-1.0704965	-6.2073803	4.0663872	0.9992852
7-3	-13.2203023	-17.7762882	-8.6643163	0.0000000
8-3	-6.4493557	-12.2134421	-0.6852694	0.0155933
9-3	4.9958200	0.5417953	9.4498448	0.0151485
5-4	2.2262360	-2.3583508	6.8108228	0.8484271
6-4	-3.5413497	-8.4656744	1.3829751	0.3800009
7-4	-15.6911554	-20.0060607	-11.3762501	0.0000000
8-4	-8.9202089	-14.4956991	-3.3447187	0.0000316
9-4	2.5249669	-1.6821385	6.7320723	0.6336442
6-5	-5.7675857	-10.8158876	-0.7192837	0.0121381
7-5	-17.9173914	-22.3732624	-13.4615204	0.0000000
8-5	-11.1464449	-16.8317302	-5.4611595	0.0000001
9-5	0.2987309	-4.0528337	4.6502955	0.9999999
7-6	-12.1498057	-16.9545246	-7.3450868	0.0000000
8-6	-5.3788592	-11.3414935	0.5837750	0.1146230
9-6	6.0663166	1.3581691	10.7744640	0.0022588
8-7	6.7709465	1.3008057	12.2410873	0.0041512
9-7	18.2161223	14.1496637	22.2825809	0.0000000
9-8	11.4451758	6.0596610	16.8306905	0.0000000

NULL

Week 5

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	14020.28	8	1752.53	38.60	< .001
Residuals	18432.48	406	45.40
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 406) = 38.60, p < .001; Eta2 = 0.43, 95% CI [0.37, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.1027092	-0.6382959	6.8437142	0.1954952
3-1	6.5315699	2.0636812	10.9994587	0.0002331
4-1	7.9987433	3.7749463	12.2225404	0.0000003
5-1	10.2344234	5.8678466	14.6010003	0.0000000
6-1	3.9636632	-0.7556378	8.6829642	0.1818097
7-1	-7.7011602	-11.7860199	-3.6163005	0.0000003
8-1	-2.1248708	-7.1934272	2.9436857	0.9288607
9-1	10.8635888	6.8918482	14.8353295	0.0000000
3-2	3.4288607	-0.8020651	7.6597866	0.2219578
4-2	4.8960341	0.9237344	8.8683338	0.0044129
5-2	7.1317142	3.0079178	11.2555106	0.0000042
6-2	0.8609540	-3.6346556	5.3565636	0.9996168
7-2	-10.8038694	-14.6281055	-6.9796332	0.0000000
8-2	-5.2275800	-10.0885437	-0.3666163	0.0243328
9-2	7.7608796	4.0577151	11.4640441	0.0000000
4-3	1.4671734	-3.1960954	6.1304422	0.9874235
5-3	3.7028535	-1.0901214	8.4958283	0.2816361
6-3	-2.5679068	-7.6842938	2.5484803	0.8230979
7-3	-14.2327301	-18.7705372	-9.6949230	0.0000000
8-3	-8.6564407	-14.0966502	-3.2162312	0.0000360
9-3	4.3320189	-0.1042339	8.7682716	0.0617002
5-4	2.2356801	-2.3306137	6.8019740	0.8424254
6-4	-4.0350801	-8.9397563	0.8695960	0.2045592
7-4	-15.6999035	-19.9975918	-11.4022152	0.0000000
8-4	-10.1236141	-15.3652084	-4.8820198	0.0000001
9-4	2.8648455	-1.3254731	7.0551641	0.4528231
6-5	-6.2707603	-11.2989190	-1.2426016	0.0037164
7-5	-17.9355836	-22.3736752	-13.4974921	0.0000000
8-5	-12.3592942	-17.7166109	-7.0019775	0.0000000
9-5	0.6291654	-3.7050360	4.9633668	0.9999527
7-6	-11.6648233	-16.4503709	-6.8792758	0.0000000
8-6	-6.0885339	-11.7370427	-0.4400252	0.0237362
9-6	6.8999256	2.2105642	11.5892871	0.0002042
8-7	5.5762894	0.4459943	10.7065845	0.0216716
9-7	18.5647490	14.5145160	22.6149820	0.0000000
9-8	12.9884596	7.9477679	18.0291513	0.0000000

NULL

Week 6

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	14704.30	8	1838.04	37.85	< .001
Residuals	20201.26	416	48.56
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 416) = 37.85, p < .001; Eta2 = 0.42, 95% CI [0.36, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.0396538	-0.8288704	6.9081780	0.2599408
3-1	5.6059359	0.9857508	10.2261210	0.0055074
4-1	7.9345270	3.5667539	12.3023000	0.0000010
5-1	9.9104184	5.3949987	14.4258382	0.0000000
6-1	5.1792367	0.2990695	10.0594039	0.0279869
7-1	-7.9437235	-12.1678232	-3.7196237	0.0000003
8-1	-1.7199713	-6.4184099	2.9784673	0.9674920
9-1	11.4551347	7.3480098	15.5622595	0.0000000
3-2	2.5662821	-1.8088627	6.9414269	0.6628087
4-2	4.8948732	0.7871702	9.0025761	0.0070550
5-2	6.8707646	2.6064009	11.1351283	0.0000264
6-2	2.1395829	-2.5092679	6.7884337	0.8836844
7-2	-10.9833773	-14.9379696	-7.0287849	0.0000000
8-2	-4.7596251	-9.2173271	-0.3019231	0.0262610
9-2	8.4154808	4.5860871	12.2448746	0.0000000
4-3	2.3285911	-2.4936339	7.1508161	0.8525849
5-3	4.3044825	-0.6518698	9.2608348	0.1480962
6-3	-0.4266992	-5.7174878	4.8640894	0.9999995
7-3	-13.5496594	-18.2421461	-8.8571726	0.0000000
8-3	-7.3259072	-12.4495522	-2.2022622	0.0003610
9-3	5.8491988	1.2617281	10.4366694	0.0026550
5-4	1.9758914	-2.7460530	6.6978359	0.9296484
6-4	-2.7552903	-7.8271514	2.3165709	0.7500831
7-4	-15.8782504	-20.3224335	-11.4340674	0.0000000
8-4	-9.6544983	-14.5517494	-4.7572471	0.0000001
9-4	3.5206077	-0.8125457	7.8537611	0.2190530
6-5	-4.7311817	-9.9307346	0.4683711	0.1077939
7-5	-17.8541419	-22.4435140	-13.2647697	0.0000000
8-5	-11.6303897	-16.6597677	-6.6010117	0.0000000
9-5	1.5447162	-2.9372245	6.0266570	0.9775410
7-6	-13.1229602	-18.0716320	-8.1742883	0.0000000
8-6	-6.8992080	-12.2584672	-1.5399488	0.0022848
9-6	6.2758980	1.4266909	11.1251050	0.0020998
8-7	6.2237522	1.4541981	10.9933063	0.0018396
9-7	19.3988581	15.2105654	23.5871508	0.0000000
9-8	13.1751059	8.5088331	17.8413788	0.0000000

NULL

Week 7

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	13588.29	8	1698.54	34.85	< .001
Residuals	20325.59	417	48.74
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 417) = 34.85, p < .001; Eta2 = 0.40, 95% CI [0.34, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.3580735	-0.5176316	7.2337786	0.1503388
3-1	5.9399863	1.3112251	10.5687475	0.0024010
4-1	8.2089811	3.8331005	12.5848618	0.0000004
5-1	10.1550979	5.6312965	14.6788994	0.0000000
6-1	5.2564617	0.3672358	10.1456876	0.0244005
7-1	-7.0264696	-11.2584102	-2.7945289	0.0000124
8-1	1.0840395	-3.5829968	5.7510757	0.9984352
9-1	12.0366852	7.9219366	16.1514338	0.0000000
3-2	2.5819128	-1.8013533	6.9651789	0.6575928
4-2	4.8509076	0.7355798	8.9662354	0.0081518
5-2	6.7970244	2.5247451	11.0693038	0.0000359
6-2	1.8983882	-2.7590920	6.5558683	0.9392081
7-2	-10.3845431	-14.3464761	-6.4226101	0.0000000
8-2	-2.2740341	-6.6976998	2.1496317	0.8029714
9-2	8.6786117	4.8421097	12.5151137	0.0000000
4-3	2.2689948	-2.5621814	7.1001710	0.8712867
5-3	4.2151116	-0.7504408	9.1806641	0.1707115
6-3	-0.6835246	-5.9841342	4.6170849	0.9999811
7-3	-12.9664559	-17.6676529	-8.2652588	0.0000000
8-3	-4.8559468	-9.9523337	0.2404400	0.0757432
9-3	6.0966989	1.5007128	10.6926850	0.0014021
5-4	1.9461168	-2.7845927	6.6768263	0.9359874
6-4	-2.9525194	-8.0337952	2.1287563	0.6743579
7-4	-15.2354507	-19.6878831	-10.7830182	0.0000000
8-4	-7.1249417	-11.9928016	-2.2570817	0.0002265
9-4	3.8277041	-0.5134927	8.1689009	0.1341496
6-5	-4.8986363	-10.1078407	0.3105681	0.0841124
7-5	-17.1815675	-21.7794586	-12.5836764	0.0000000
8-5	-9.0710585	-14.0723091	-4.0698079	0.0000010
9-5	1.8815872	-2.6086730	6.3718475	0.9291091
7-6	-12.2829312	-17.2407890	-7.3250735	0.0000000
8-6	-4.1724222	-9.5064881	1.1616436	0.2656625
9-6	6.7802235	1.9220152	11.6384318	0.0005719
8-7	8.1105090	3.3716220	12.8493961	0.0000055
9-7	19.0631548	14.8670876	23.2592219	0.0000000
9-8	10.9526457	6.3181139	15.5871775	0.0000000

NULL

Week 8

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	12626.16	8	1578.27	34.52	< .001
Residuals	18929.53	414	45.72
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 414) = 34.52, p < .001; Eta2 = 0.40, 95% CI [0.34, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.0546879	-0.6992193	6.8085951	0.2172327
3-1	6.4928132	1.9335804	11.0520460	0.0003905
4-1	8.3427710	4.1044067	12.5811353	0.0000001
5-1	9.3744503	4.9928138	13.7560868	0.0000000
6-1	5.6876364	0.9018493	10.4734235	0.0073145
7-1	-6.7066213	-10.8055691	-2.6076736	0.0000181
8-1	1.8263438	-2.6940262	6.3467138	0.9421330
9-1	11.6880541	7.7026155	15.6734927	0.0000000
3-2	3.4381253	-0.8875035	7.7637542	0.2455252
4-2	5.2880832	1.3020836	9.2740827	0.0014005
5-2	6.3197624	2.1817437	10.4577812	0.0000919
6-2	2.6329485	-1.9308458	7.1967428	0.6829155
7-2	-9.7613092	-13.5987346	-5.9238838	0.0000000
8-2	-1.2283441	-5.5129917	3.0563036	0.9932178
9-2	8.6333662	4.9174301	12.3493024	0.0000000
4-3	1.8499578	-2.9021972	6.6021128	0.9532207
5-3	2.8816371	-1.9987301	7.7620043	0.6545483
6-3	-0.8051768	-6.0514139	4.4410602	0.9999278
7-3	-13.1994345	-17.8276759	-8.5711932	0.0000000
8-3	-4.6664694	-9.6717656	0.3388268	0.0899318
9-3	5.1952409	0.6672208	9.7232610	0.0115080
5-4	1.0316793	-3.5503630	5.6137216	0.9987483
6-4	-2.6551347	-7.6250573	2.3147880	0.7669725
7-4	-15.0493924	-19.3619028	-10.7368820	0.0000000
8-4	-6.5164272	-11.2313098	-1.8015446	0.0006846
9-4	3.3452830	-0.8594873	7.5500534	0.2443358
6-5	-3.6868139	-8.7794692	1.4058413	0.3707210
7-5	-16.0810716	-20.5344695	-11.6276738	0.0000000
8-5	-7.5481065	-12.3921879	-2.7040251	0.0000585
9-5	2.3136038	-2.0355456	6.6627532	0.7711615
7-6	-12.3942577	-17.2458319	-7.5426835	0.0000000
8-6	-3.8612925	-9.0737914	1.3512063	0.3384645
9-6	6.0004177	1.2443563	10.7564791	0.0031223
8-7	8.5329652	3.9430023	13.1229280	0.0000005
9-7	18.3946754	14.3304738	22.4588770	0.0000000
9-8	9.8617103	5.3728232	14.3505973	0.0000000

NULL

Week 9

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	11792.01	7	1684.57	39.07	< .001
Residuals	16297.97	378	43.12
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(7, 378) = 39.07, p < .001; Eta2 = 0.42, 95% CI [0.35, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.1172034	-0.4459342	6.680341	0.1362526
4-1	8.4669906	4.4440157	12.489966	0.0000000
5-1	9.9324155	5.7734492	14.091382	0.0000000
6-1	6.2478009	1.7052225	10.790379	0.0008945
7-1	-6.2023255	-10.1339517	-2.270699	0.0000598
8-1	2.7329487	-1.5577008	7.023598	0.5232263
9-1	11.5723402	7.7894375	15.355243	0.0000000
4-2	5.3497873	1.5663521	9.133223	0.0005480
5-2	6.8152121	2.8874832	10.742941	0.0000058
6-2	3.1305975	-1.2012695	7.462465	0.3523943
7-2	-9.3195288	-13.0056838	-5.633374	0.0000000
8-2	-0.3842547	-4.4511610	3.682652	0.9999919
9-2	8.4551369	4.9280408	11.982233	0.0000000
5-4	1.4654248	-2.8837628	5.814612	0.9701504
6-4	-2.2191898	-6.9365461	2.498166	0.8410305
7-4	-14.6693161	-18.8016420	-10.536990	0.0000000
8-4	-5.7340420	-10.2093191	-1.258765	0.0027927
9-4	3.1053496	-0.8857387	7.096438	0.2584822
6-5	-3.6846146	-8.5184664	1.149237	0.2836228
7-5	-16.1347409	-20.3995727	-11.869909	0.0000000
8-5	-7.1994668	-11.7973770	-2.601557	0.0000706
9-5	1.6399248	-2.4882053	5.768055	0.9284689
7-6	-12.4501263	-17.0898254	-7.810427	0.0000000
8-6	-3.5148522	-8.4624572	1.432753	0.3752894
9-6	5.3245394	0.8101761	9.838903	0.0087125
8-7	8.9352742	4.5419312	13.328617	0.0000000
9-7	17.7746657	13.8756732	21.673658	0.0000000
9-8	8.8393916	4.5786251	13.100158	0.0000000

NULL

Week 10

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	7389.02	6	1231.50	27.63	< .001
Residuals	14218.88	319	44.57
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 319) = 27.63, p < .001; Eta2 = 0.34, 95% CI [0.27, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.0526806	-0.4742024	6.5795636	0.1393582
4-1	8.4287083	4.4466666	12.4107500	0.0000000
5-1	9.1425096	5.0258604	13.2591589	0.0000000
6-1	6.8534719	2.3571137	11.3498300	0.0001736
7-1	-5.8653066	-9.8234710	-1.9071422	0.0002993
8-1	2.8606948	-1.3862978	7.1076874	0.4175103
4-2	5.3760277	1.6310885	9.1209669	0.0005336
5-2	6.0898290	2.2020643	9.9775937	0.0000997
6-2	3.8007913	-0.4869995	8.0885820	0.1204344
7-2	-8.9179872	-12.6375274	-5.1984471	0.0000000
8-2	-0.1919858	-4.2175118	3.8335402	0.9999993
5-4	0.7138013	-3.5911338	5.0187365	0.9989461
6-4	-1.5752364	-6.2445942	3.0941213	0.9535304
7-4	-14.2940149	-18.4476556	-10.1403742	0.0000000
8-4	-5.5680135	-9.9977552	-1.1382717	0.0042188
6-5	-2.2890378	-7.0737057	2.4956301	0.7910005
7-5	-15.0078162	-19.2906746	-10.7249578	0.0000000
8-5	-6.2818148	-10.8329418	-1.7306878	0.0010408
7-6	-12.7187785	-17.3677903	-8.0697666	0.0000000
8-6	-3.9927770	-8.8900408	0.9044867	0.1939293
8-7	8.7260014	4.3177114	13.1342915	0.0000002

NULL

Week 11

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	5404.25	5	1080.85	23.11	< .001
Residuals	13422.65	287	46.77
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 287) = 23.11, p < .001; Eta2 = 0.29, 95% CI [0.21, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6244016	-0.8687566	6.1175599	0.2621458
4-1	7.3495216	3.4055570	11.2934863	0.0000027
5-1	8.1265423	4.0492572	12.2038273	0.0000004
7-1	-5.4011882	-9.3952803	-1.4070960	0.0017906
8-1	3.6952774	-0.5111047	7.9016594	0.1215739
4-2	4.7251200	1.0159907	8.4342493	0.0041065
5-2	5.5021407	1.6515515	9.3527298	0.0007601
7-2	-8.0255898	-11.7879767	-4.2632029	0.0000000
8-2	1.0708757	-2.9161574	5.0579089	0.9722371
5-4	0.7770207	-3.4867499	5.0407912	0.9952430
7-4	-12.7507098	-16.9349968	-8.5664228	0.0000000
8-4	-3.6542443	-8.0416280	0.7331394	0.1633060
7-5	-13.5277305	-17.8379107	-9.2175502	0.0000000
8-5	-4.4312649	-8.9388732	0.0763433	0.0570822
8-7	9.0964655	4.6639662	13.5289649	0.0000002

NULL

Week 12

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	5308.19	5	1061.64	24.60	< .001
Residuals	12212.13	283	43.15
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 283) = 24.60, p < .001; Eta2 = 0.30, 95% CI [0.23, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.5850866	-0.7706211	5.9407944	0.2361785
4-1	7.9012725	4.1124970	11.6900480	0.0000001
5-1	7.7964830	3.8796330	11.7133330	0.0000004
7-1	-5.4260186	-9.3718340	-1.4802033	0.0013955
8-1	4.1486809	0.1078137	8.1895481	0.0403625
4-2	5.3161859	1.7530052	8.8793665	0.0003643
5-2	5.2113964	1.5123221	8.9104706	0.0009571
7-2	-8.0111053	-11.7408365	-4.2813741	0.0000000
8-2	1.5635943	-2.2665551	5.3937436	0.8501596
5-4	-0.1047895	-4.2007871	3.9912080	0.9999997
7-4	-13.3272912	-17.4509959	-9.2035864	0.0000000
8-4	-3.7525916	-7.9673383	0.4621551	0.1121841
7-5	-13.2225016	-17.4641799	-8.9808234	0.0000000
8-5	-3.6478021	-7.9780427	0.6824385	0.1538091
8-7	9.5746996	5.2182413	13.9311578	0.0000000

NULL

Week 13

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	4585.50	5	917.10	21.15	< .001
Residuals	12183.21	281	43.36
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 281) = 21.15, p < .001; Eta2 = 0.27, 95% CI [0.20, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.3774645	0.0136611	6.7412679	0.0484177
4-1	8.1710155	4.3730996	11.9689315	0.0000000
5-1	7.6982301	3.7719307	11.6245295	0.0000007
7-1	-4.1289287	-8.1463249	-0.1115326	0.0400092
8-1	4.9892479	0.9386321	9.0398637	0.0062980
4-2	4.7935510	1.2217742	8.3653278	0.0020048
5-2	4.3207656	0.6127673	8.0287639	0.0119677
7-2	-7.5063933	-11.3107192	-3.7020673	0.0000006
8-2	1.6117834	-2.2276062	5.4511730	0.8344079
5-4	-0.4727854	-4.5786645	3.6330938	0.9994745
7-4	-12.2999443	-16.4930205	-8.1068680	0.0000000
8-4	-3.1817676	-7.4066824	1.0431472	0.2595433
7-5	-11.8271589	-16.1368630	-7.5174547	0.0000000
8-5	-2.7089822	-7.0496695	1.6317051	0.4733373
8-7	9.1181766	4.6949187	13.5414346	0.0000001

NULL

Week 14

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	4159.97	5	831.99	18.45	< .001
Residuals	12672.91	281	45.10
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 281) = 18.45, p < .001; Eta2 = 0.25, 95% CI [0.17, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.141740	-0.2890009	6.5724806	0.0940028
4-1	7.861652	3.9881604	11.7351441	0.0000002
5-1	6.856717	2.8522870	10.8611470	0.0000224
7-1	-3.849219	-7.9465589	0.2481202	0.0792675
8-1	5.811176	1.6799556	9.9423961	0.0009820
4-2	4.719912	1.0770596	8.3627651	0.0032826
5-2	3.714977	-0.0668077	7.4967620	0.0574025
7-2	-6.990959	-10.8709886	-3.1109298	0.0000066
8-2	2.669436	-1.2463548	6.5852268	0.3704026
5-4	-1.004935	-5.1925185	3.1826481	0.9831156
7-4	-11.710872	-15.9873871	-7.4343560	0.0000000
8-4	-2.050476	-6.3594640	2.2585113	0.7476305
7-5	-10.705936	-15.1014006	-6.3104721	0.0000000
8-5	-1.045541	-5.4726051	3.3815228	0.9842862
8-7	9.660395	5.1491175	14.1716730	0.0000000

NULL

Week 15

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1718.28	4	429.57	9.60	< .001
Residuals	10965.29	245	44.76
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 9.60, p < .001; Eta2 = 0.14, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6307040	-0.6425016	5.903910	0.1798951
4-1	7.3674144	3.6717881	11.063041	0.0000011
5-1	5.9265812	2.1060292	9.747133	0.0002768
8-1	5.1875960	1.2460759	9.129116	0.0033028
4-2	4.7367103	1.2611326	8.212288	0.0020779
5-2	3.2958771	-0.3122532	6.904007	0.0916338
8-2	2.5568920	-1.1790909	6.292875	0.3304011
5-4	-1.4408332	-5.4361282	2.554462	0.8591905
8-4	-2.1798183	-6.2909430	1.931306	0.5913006
8-5	-0.7389852	-4.9627642	3.484794	0.9890488

NULL

Week 16

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1842.98	4	460.75	9.87	< .001
Residuals	11441.67	245	46.70
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 9.87, p < .001; Eta2 = 0.14, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6584406	-0.6851095	6.001991	0.1888063
4-1	7.4133038	3.6382548	11.188353	0.0000016
5-1	6.3856509	2.4829914	10.288310	0.0001036
8-1	5.4814808	1.4552535	9.507708	0.0021077
4-2	4.7548632	1.2045917	8.305135	0.0026299
5-2	3.7272103	0.0415376	7.412883	0.0459945
8-2	2.8230402	-0.9932327	6.639313	0.2534124
5-4	-1.0276529	-5.1088109	3.053505	0.9580838
8-4	-1.9318230	-6.1312999	2.267654	0.7133818
8-5	-0.9041701	-5.2187224	3.410382	0.9784851

NULL

Week 17

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1613.15	4	403.29	9.12	< .001
Residuals	10828.06	245	44.20
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 9.12, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.7500035	-0.5026557	6.002663	0.1409517
4-1	7.3442407	3.6718127	11.016669	0.0000010
5-1	5.5544634	1.7578938	9.351033	0.0007320
8-1	4.9765052	1.0597268	8.893284	0.0051157
4-2	4.5942373	1.1404764	8.047998	0.0028770
5-2	2.8044599	-0.7810215	6.389941	0.2027516
8-2	2.2265017	-1.4860297	5.939033	0.4681827
5-4	-1.7897774	-5.7599931	2.180438	0.7285893
8-4	-2.3677356	-6.4530539	1.717583	0.5034401
8-5	-0.5779582	-4.7752238	3.619307	0.9956218

NULL

Week 18

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1476.43	4	369.11	8.49	< .001
Residuals	10651.31	245	43.47
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 8.49, p < .001; Eta2 = 0.12, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.4858398	-0.7401621	5.711842	0.2158672
4-1	6.9747089	3.3323783	10.617040	0.0000031
5-1	4.7873105	1.0218559	8.552765	0.0050764
8-1	5.3034031	1.4187248	9.188081	0.0020291
4-2	4.4888691	1.0634136	7.914325	0.0034912
5-2	2.3014707	-1.2546257	5.857567	0.3884204
8-2	2.8175634	-0.8645419	6.499669	0.2221118
5-4	-2.1873984	-6.1250761	1.750279	0.5461634
8-4	-1.6713057	-5.7231427	2.380531	0.7885849
8-5	0.5160926	-3.6467742	4.678959	0.9970863

NULL

Week 19

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	440.17	3	146.72	3.16	0.026
Residuals	9336.02	201	46.45
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and small (F(3, 201) = 3.16, p = 0.026; Eta2 = 0.05, 95% CI [3.02e-03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.4311200	-0.7122075	5.574447	0.1900496
5-1	4.1678055	0.4988501	7.836761	0.0189029
8-1	2.8485942	-0.9365295	6.633718	0.2108565
5-2	1.7366855	-1.7282770	5.201648	0.5649719
8-2	0.4174742	-3.1702679	4.005216	0.9904586
8-5	-1.3192113	-5.3753942	2.736972	0.8340526

NULL

Week 20

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	350.37	3	116.79	2.53	0.059
Residuals	9288.26	201	46.21
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(3, 201) = 2.53, p = 0.059; Eta2 = 0.04, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.7050189	-0.4302579	5.840296	0.1173461
5-1	3.3804976	-0.2790610	7.040056	0.0816576
8-1	2.7502497	-1.0251796	6.525679	0.2368045
5-2	0.6754786	-2.7806095	4.131567	0.9575067
8-2	0.0452308	-3.5333223	3.623784	0.9999873
8-5	-0.6302479	-4.6760420	3.415546	0.9777188

NULL

Week 21

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	285.27	3	95.09	2.20	0.089
Residuals	8691.82	201	43.24
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(3, 201) = 2.20, p = 0.089; Eta2 = 0.03, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.4292731	-0.6036689	5.462215	0.1648439
5-1	2.8371689	-0.7029423	6.377280	0.1644339
8-1	2.8280969	-0.8241030	6.480297	0.1891544
5-2	0.4078958	-2.9353862	3.751178	0.9890394
8-2	0.3988239	-3.0629259	3.860574	0.9907331
8-5	-0.0090719	-3.9228121	3.904668	0.9999999

NULL

Week 22

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	454.98	2	227.49	5.36	0.006
Residuals	6839.21	161	42.48
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 161) = 5.36, p = 0.006; Eta2 = 0.06, 95% CI [0.01, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.621362	-0.1235423	5.366265	0.0646433
8-1	4.445097	1.1113285	7.778865	0.0054270
8-2	1.823735	-1.3392189	4.986689	0.3622754

NULL

Week 23

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	592.03	2	296.02	6.51	0.002
Residuals	7230.31	159	45.47
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 6.51, p = 0.002; Eta2 = 0.08, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.247427	0.4071025	6.087752	0.0205897
8-1	5.004214	1.4912529	8.517176	0.0026992
8-2	1.756787	-1.5827743	5.096348	0.4288207

NULL

Week 24

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	698.79	2	349.39	7.61	< .001
Residuals	7298.07	159	45.90
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 7.61, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.189458	0.3358549	6.043062	0.0243150
8-1	5.621529	2.0921443	9.150913	0.0006736
8-2	2.432071	-0.9231033	5.787244	0.2027588

NULL

Week 25

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	479.66	2	239.83	6.09	0.003
Residuals	6260.81	159	39.38
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 6.09, p = 0.003; Eta2 = 0.07, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.537799	-0.1052487	5.180847	0.0627888
8-1	4.699379	1.4304131	7.968344	0.0024340
8-2	2.161580	-0.9460294	5.269189	0.2296551

NULL

Week 26

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	732.14	2	366.07	9.98	< .001
Residuals	5834.69	159	36.70
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.98, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.983246	0.4317276	5.534764	0.0173505
8-1	5.855742	2.6999812	9.011502	0.0000609
8-2	2.872496	-0.1274960	5.872488	0.0637389

NULL

Week 27

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	611.53	2	305.76	9.48	< .001
Residuals	5128.41	159	32.25
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.48, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.791087	0.3989767	5.183197	0.0176324
8-1	5.331819	2.3732167	8.290421	0.0001015
8-2	2.540732	-0.2718333	5.353297	0.0856555

NULL

Week 28

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	601.83	2	300.92	7.67	< .001
Residuals	6238.90	159	39.24
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 7.67, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.266623	0.6282056	5.905041	0.0108253
8-1	5.050466	1.7872272	8.313705	0.0009889
8-2	1.783843	-1.3183222	4.886008	0.3641931

NULL

Week 29

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	902.92	2	451.46	12.29	< .001
Residuals	5842.97	159	36.75
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 12.29, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.618120	1.0647923	6.171448	0.0028626
8-1	6.393294	3.2352952	9.551292	0.0000113
8-2	2.775174	-0.2269458	5.777293	0.0764988

NULL

Week 30

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	790.30	2	395.15	11.32	< .001
Residuals	5550.74	159	34.91
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 11.32, p < .001; Eta2 = 0.12, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.545213	1.0565552	6.033871	0.0026981
8-1	5.904671	2.8266571	8.982685	0.0000330
8-2	2.359458	-0.5666251	5.285541	0.1398156

NULL

Week 31

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	599.56	2	299.78	8.37	< .001
Residuals	5695.93	159	35.82
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 8.37, p < .001; Eta2 = 0.10, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.905607	0.3846111	5.426603	0.0193770
8-1	5.227656	2.1096460	8.345666	0.0003226
8-2	2.322049	-0.6420557	5.286154	0.1558365

NULL

Week 32

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	514.19	2	257.09	6.65	0.002
Residuals	6145.48	159	38.65
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 6.65, p = 0.002; Eta2 = 0.08, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.176745	0.5581539	5.795337	0.0128761
8-1	4.551938	1.3132205	7.790656	0.0031331
8-2	1.375193	-1.7036614	4.454047	0.5423411

NULL

Week 33

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	767.98	2	383.99	10.67	< .001
Residuals	5724.21	159	36.00
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 10.67, p < .001; Eta2 = 0.12, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.581162	1.0539173	6.108407	0.0028626
8-1	5.772860	2.6471208	8.898599	0.0000663
8-2	2.191698	-0.7797548	5.163150	0.1918014

NULL

Week 34

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	989.37	2	494.69	12.38	< .001
Residuals	6314.03	158	39.96
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 12.38, p < .001; Eta2 = 0.14, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	4.308961	1.638395	6.979528	0.0005637
8-1	6.395990	3.102593	9.689388	0.0000261
8-2	2.087029	-1.050413	5.224471	0.2599581

NULL

Week 35

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	834.68	2	417.34	9.79	< .001
Residuals	6736.35	158	42.64
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 9.79, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.944952	1.186519	6.703386	0.0025833
8-1	5.884152	2.482394	9.285910	0.0001996
8-2	1.939200	-1.301471	5.179870	0.3352120

NULL

Week 36

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	854.78	2	427.39	9.59	< .001
Residuals	6998.03	157	44.57
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 157) = 9.59, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.395902	0.5604359	6.231368	0.0143255
8-1	6.307131	2.8166386	9.797623	0.0000974
8-2	2.911229	-0.4024883	6.224946	0.0975533

NULL

Week 37

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	843.12	1	843.12	16.28	< .001
Residuals	4349.28	84	51.78
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 84) = 16.28, p < .001; Eta2 = 0.16, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	6.438708	3.265687	9.611729	0.0001198

NULL

Week 38

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	966.36	1	966.36	17.47	< .001
Residuals	4590.70	83	55.31
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 83) = 17.47, p < .001; Eta2 = 0.17, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	6.918628	3.626494	10.21076	7.18e-05

NULL

Week 39

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	881.83	1	881.83	16.28	< .001
Residuals	4387.66	81	54.17
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 81) = 16.28, p < .001; Eta2 = 0.17, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	6.696864	3.394404	9.999325	0.0001231

NULL

Week 40

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1079.68	1	1079.68	20.13	< .001
Residuals	4291.12	80	53.64
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 80) = 20.13, p < .001; Eta2 = 0.20, 95% CI [0.09, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	7.438654	4.139108	10.7382	2.4e-05

NULL

Week 41

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	759.58	1	759.58	12.15	< .001
Residuals	3937.36	63	62.50
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 63) = 12.15, p < .001; Eta2 = 0.16, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	6.83773	2.918268	10.75719	0.0008981

NULL

Week 42

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	604.37	1	604.37	9.62	0.003
Residuals	3770.81	60	62.85
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(1, 60) = 9.62, p = 0.003; Eta2 = 0.14, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	6.247564	2.217644	10.27748	0.0029368

NULL

Week 43

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	472.47	1	472.47	8.28	0.006
Residuals	3195.64	56	57.07
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(1, 56) = 8.28, p = 0.006; Eta2 = 0.13, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	5.721866	1.738323	9.705409	0.0056649

NULL

Model fitting for DMI (kg) as output variable

Week 1

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1555.18	6	259.20	22.11	< .001
Residuals	3963.20	338	11.73
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 338) = 22.11, p < .001; Eta2 = 0.28, 95% CI [0.21, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.4386388	-1.3696248	2.2469024	0.9913566
3-1	-2.7939389	-4.9535512	-0.6343265	0.0028059
5-1	1.1184541	-0.9921878	3.2290960	0.7004593
6-1	-2.2641282	-4.5452639	0.0170075	0.0532000
7-1	-5.4004388	-7.3749090	-3.4259686	0.0000000
9-1	-0.4244932	-2.3442859	1.4952994	0.9947599
3-2	-3.2325777	-5.2776509	-1.1875045	0.0000813
5-2	0.6798153	-1.3134755	2.6731061	0.9511589
6-2	-2.7027670	-4.8757785	-0.5297555	0.0048460
7-2	-5.8390776	-7.6875721	-3.9905832	0.0000000
9-2	-0.8631320	-2.6531049	0.9268408	0.7850805
5-3	3.9123930	1.5956460	6.2291399	0.0000182
6-3	0.5298107	-1.9432617	3.0028831	0.9955923
7-3	-2.6064999	-4.7999082	-0.4130917	0.0086611
9-3	2.3694456	0.2251250	4.5137663	0.0196434
6-5	-3.3825823	-5.8130084	-0.9521562	0.0009017
7-5	-6.5188929	-8.6641023	-4.3736834	0.0000000
9-5	-1.5429473	-3.6379401	0.5520455	0.3066354
7-6	-3.1363106	-5.4494675	-0.8231538	0.0013794
9-6	1.8396349	-0.4270291	4.1062990	0.1986389
9-7	4.9759456	3.0182126	6.9336785	0.0000000

NULL

Week 2

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2388.08	6	398.01	31.71	< .001
Residuals	4242.40	338	12.55
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 338) = 31.71, p < .001; Eta2 = 0.36, 95% CI [0.29, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1651730	-1.7057019	2.0360480	0.9999732
3-1	-1.1532886	-3.3876778	1.0811006	0.7259378
5-1	2.5918899	0.4081668	4.7756131	0.0087876
6-1	-0.0457822	-2.4059025	2.3143382	1.0000000
7-1	-6.5120531	-8.5548896	-4.4692166	0.0000000
9-1	0.8150945	-1.1711712	2.8013603	0.8872075
3-2	-1.3184617	-3.4343458	0.7974225	0.5163292
5-2	2.4267169	0.3644081	4.4890256	0.0097483
6-2	-0.2109552	-2.4592075	2.0372972	0.9999616
7-2	-6.6772261	-8.5897249	-4.7647273	0.0000000
9-2	0.6499215	-1.2020295	2.5018724	0.9440971
5-3	3.7451785	1.3482140	6.1421431	0.0001037
6-3	1.1075065	-1.4511964	3.6662094	0.8590427
7-3	-5.3587645	-7.6281198	-3.0894092	0.0000000
9-3	1.9683831	-0.2501849	4.1869511	0.1198178
6-5	-2.6376721	-5.1522520	-0.1230922	0.0327966
7-5	-9.1039430	-11.3234306	-6.8844554	0.0000000
9-5	-1.7767954	-3.9443276	0.3907368	0.1887447
7-6	-6.4662709	-8.8595211	-4.0730207	0.0000000
9-6	0.8608767	-1.4842709	3.2060242	0.9311307
9-7	7.3271476	5.3016279	9.3526673	0.0000000

NULL

Week 3

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3158.08	8	394.76	34.78	< .001
Residuals	4358.26	384	11.35
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 384) = 34.78, p < .001; Eta2 = 0.42, 95% CI [0.35, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.4951625	-1.3758876	2.3662126	0.9960576
3-1	0.2237412	-2.0108572	2.4583395	0.9999973
4-1	4.4715469	2.3590302	6.5840636	0.0000000
5-1	3.0300447	0.8461172	5.2139722	0.0006436
6-1	0.6243400	-1.7360012	2.9846812	0.9960711
7-1	-5.8669773	-7.9100050	-3.8239496	0.0000000
8-1	-2.8652660	-9.0962951	3.3657632	0.8838760
9-1	1.6484411	-0.3380105	3.6348928	0.1947262
3-2	-0.2714214	-2.3875035	1.8446608	0.9999817
4-2	3.9763844	1.9896531	5.9631156	0.0000000
5-2	2.5348822	0.4723804	4.5973839	0.0046110
6-2	0.1291775	-2.1192853	2.3776402	1.0000000
7-2	-6.3621398	-8.2748176	-4.4494620	0.0000000
8-2	-3.3604285	-9.5499437	2.8290867	0.7502142
9-2	1.1532786	-0.6988456	3.0054029	0.5845749
4-3	4.2478057	1.9154888	6.5801227	0.0000009
5-3	2.8063035	0.4091146	5.2034925	0.0089418
6-3	0.4005988	-2.1583435	2.9595412	0.9999156
7-3	-6.0907185	-8.3602862	-3.8211508	0.0000000
8-3	-3.0890071	-9.3979452	3.2199310	0.8421224
9-3	1.4247000	-0.7940757	3.6434757	0.5421283
5-4	-1.4415022	-3.7253175	0.8423130	0.5660113
6-4	-3.8472069	-6.3002628	-1.3941510	0.0000512
7-4	-10.3385242	-12.4879974	-8.1890511	0.0000000
8-4	-7.3368129	-13.6035502	-1.0700755	0.0089328
9-4	-2.8231057	-4.9188784	-0.7273331	0.0010880
6-5	-2.4057047	-4.9205200	0.1091106	0.0733018
7-5	-8.8970220	-11.1167173	-6.6773267	0.0000000
8-5	-5.8953106	-12.1864798	0.3958585	0.0864065
9-5	-1.3816035	-3.5493386	0.7861315	0.5525327
7-6	-6.4913173	-8.8847915	-4.0978431	0.0000000
8-6	-3.4896060	-9.8441697	2.8649578	0.7382417
9-6	1.0241012	-1.3212659	3.3694682	0.9111818
8-7	3.0017114	-3.2419440	9.2453667	0.8553170
9-7	7.5154185	5.4897092	9.5411278	0.0000000
9-8	4.5137071	-1.7116652	10.7390794	0.3681595

NULL

Week 4

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3621.69	8	452.71	41.19	< .001
Residuals	4418.39	402	10.99
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 402) = 41.19, p < .001; Eta2 = 0.45, 95% CI [0.39, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.9198041	-0.9209764	2.7605846	0.8266208
3-1	1.3264039	-0.8720435	3.5248512	0.6268353
4-1	5.8892756	3.8109349	7.9676164	0.0000000
5-1	3.3904937	1.2418974	5.5390899	0.0000439
6-1	1.1204297	-1.2017263	3.4425856	0.8530036
7-1	-5.2450386	-7.2550145	-3.2350626	0.0000000
8-1	-1.7789716	-4.4312564	0.8733132	0.4803515
9-1	2.3777621	0.4234469	4.3320772	0.0053093
3-2	0.4065997	-1.6752488	2.4884482	0.9995565
4-2	4.9694715	3.0148813	6.9240617	0.0000000
5-2	2.4706895	0.4415546	4.4998244	0.0052523
6-2	0.2006255	-2.0114620	2.4127130	0.9999988
7-2	-6.1648427	-8.0465775	-4.2831079	0.0000000
8-2	-2.6987757	-5.2552454	-0.1423060	0.0295795
9-2	1.4579579	-0.3642029	3.2801188	0.2372693
4-3	4.5628718	2.2682867	6.8574568	0.0000000
5-3	2.0640898	-0.2943178	4.4224974	0.1408223
6-3	-0.2059742	-2.7235184	2.3115700	0.9999995
7-3	-6.5714424	-8.8042934	-4.3385914	0.0000000
8-3	-3.1053755	-5.9303064	-0.2804445	0.0191121
9-3	1.0513582	-1.1315224	3.2342389	0.8542572
5-4	-2.4987819	-4.7456500	-0.2519139	0.0167359
6-4	-4.7688460	-7.1822167	-2.3554752	0.0000001
7-4	-11.1343142	-13.2490135	-9.0196148	0.0000000
8-4	-7.6682472	-10.4007488	-4.9357456	0.0000000
9-4	-3.5115135	-5.5733811	-1.4496460	0.0000064
6-5	-2.2700640	-4.7441950	0.2040670	0.1014109
7-5	-8.6355322	-10.8193177	-6.4517468	0.0000000
8-5	-5.1694653	-7.9557765	-2.3831540	0.0000005
9-5	-1.0127316	-3.1453974	1.1199342	0.8639427
7-6	-6.3654682	-8.7202211	-4.0107153	0.0000000
8-6	-2.8994013	-5.8216389	0.0228364	0.0536621
9-6	1.2573324	-1.0500916	3.5647565	0.7467111
8-7	3.4660670	0.7851963	6.1469377	0.0021395
9-7	7.6228006	5.6298629	9.6157383	0.0000000
9-8	4.1567337	1.5173375	6.7961298	0.0000460

NULL

Week 5

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3442.66	8	430.33	37.11	< .001
Residuals	4708.51	406	11.60
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 406) = 37.11, p < .001; Eta2 = 0.42, 95% CI [0.36, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.4346908	-0.4560767	3.3254583	0.3055102
3-1	2.0159897	-0.2421572	4.2741366	0.1232601
4-1	5.7765166	3.6417378	7.9112954	0.0000000
5-1	3.7020585	1.4951164	5.9090007	0.0000096
6-1	2.6210650	0.2358501	5.0062799	0.0191865
7-1	-4.8250105	-6.8895680	-2.7604530	0.0000000
8-1	-0.7769949	-3.3387294	1.7847397	0.9901084
9-1	2.7680335	0.7606483	4.7754187	0.0007153
3-2	0.5812989	-1.5570829	2.7196807	0.9952721
4-2	4.3418258	2.3341581	6.3494936	0.0000000
5-2	2.2673678	0.1831310	4.3516045	0.0214680
6-2	1.1863742	-1.0857832	3.4585317	0.7887722
7-2	-6.2597013	-8.1925352	-4.3268674	0.0000000
8-2	-2.2116856	-4.6684993	0.2451280	0.1163232
9-2	1.3333427	-0.5382995	3.2049850	0.3935454
4-3	3.7605269	1.4036316	6.1174222	0.0000337
5-3	1.6860689	-0.7363820	4.1085198	0.4269913
6-3	0.6050753	-1.9808336	3.1909843	0.9983481
7-3	-6.8410002	-9.1344850	-4.5475154	0.0000000
8-3	-2.7929845	-5.5425588	-0.0434103	0.0432463
9-3	0.7520438	-1.4901137	2.9942013	0.9810419
5-4	-2.0744581	-4.3823405	0.2334244	0.1175711
6-4	-3.1554516	-5.6343582	-0.6765449	0.0027284
7-4	-10.6015271	-12.7736518	-8.4294024	0.0000000
8-4	-6.5535115	-9.2027023	-3.9043207	0.0000000
9-4	-3.0084831	-5.1263413	-0.8906249	0.0004108
6-5	-1.0809935	-3.6223104	1.4603233	0.9229836
7-5	-8.5270691	-10.7701559	-6.2839822	0.0000000
8-5	-4.4790534	-7.1867323	-1.7713745	0.0000137
9-5	-0.9340250	-3.1246041	1.2565540	0.9219820
7-6	-7.4460755	-9.8647726	-5.0273785	0.0000000
8-6	-3.3980599	-6.2529122	-0.5432076	0.0071705
9-6	0.1469685	-2.2231144	2.5170514	0.9999999
8-7	4.0480156	1.4550773	6.6409540	0.0000562
9-7	7.5930440	5.5459875	9.6401006	0.0000000
9-8	3.5450284	0.9973771	6.0926796	0.0006060

NULL

Week 6

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3740.21	8	467.53	37.79	< .001
Residuals	5146.15	416	12.37
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 416) = 37.79, p < .001; Eta2 = 0.42, 95% CI [0.36, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.5750734	-0.3774550	3.5276017	0.2275336
3-1	2.8086231	0.4767151	5.1405311	0.0061010
4-1	6.0244127	3.8199025	8.2289229	0.0000000
5-1	3.6979196	1.4188890	5.9769503	0.0000222
6-1	3.2431594	0.7800328	5.7062860	0.0015876
7-1	-4.9556088	-7.0876039	-2.8236137	0.0000000
8-1	-0.4219460	-2.7933502	1.9494583	0.9997790
9-1	2.7533232	0.6803678	4.8262785	0.0013728
3-2	1.2335497	-0.9746811	3.4417806	0.7201486
4-2	4.4493393	2.3760922	6.5225865	0.0000000
5-2	2.1228463	-0.0294709	4.2751635	0.0565675
6-2	1.6680860	-0.6782902	4.0144623	0.3966424
7-2	-6.5306822	-8.5266510	-4.5347133	0.0000000
8-2	-1.9970193	-4.2469186	0.2528799	0.1282732
9-2	1.1782498	-0.7545286	3.1110282	0.6135165
4-3	3.2157896	0.7819076	5.6496716	0.0014988
5-3	0.8892966	-1.6122824	3.3908755	0.9728011
6-3	0.4345363	-2.2358399	3.1049125	0.9998873
7-3	-7.7642319	-10.1326321	-5.3958317	0.0000000
8-3	-3.2305691	-5.8165842	-0.6445539	0.0036193
9-3	-0.0552999	-2.3706962	2.2600964	1.0000000
5-4	-2.3264931	-4.7097612	0.0567751	0.0618821
6-4	-2.7812533	-5.3411321	-0.2213745	0.0217642
7-4	-10.9800215	-13.2230974	-8.7369456	0.0000000
8-4	-6.4463587	-8.9181080	-3.9746094	0.0000000
9-4	-3.2710895	-5.4581264	-1.0840526	0.0001444
6-5	-0.4547602	-3.0790878	2.1695673	0.9998188
7-5	-8.6535284	-10.9698845	-6.3371724	0.0000000
8-5	-4.1198656	-6.6583022	-1.5814290	0.0000221
9-5	-0.9445965	-3.2067295	1.3175366	0.9304587
7-6	-8.1987682	-10.6964706	-5.7010658	0.0000000
8-6	-3.6651054	-6.3700402	-0.9601706	0.0009749
9-6	-0.4898362	-2.9373366	1.9576641	0.9994700
8-7	4.5336628	2.1263650	6.9409606	0.0000003
9-7	7.7089320	5.5950094	9.8228545	0.0000000
9-8	3.1752692	0.8200997	5.5304387	0.0010636

NULL

Week 7

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3741.17	8	467.65	39.05	< .001
Residuals	4993.56	417	11.97
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 417) = 39.05, p < .001; Eta2 = 0.43, 95% CI [0.37, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.7824324	-0.1385974	3.7034623	0.0932859
3-1	3.0730803	0.7787910	5.3673695	0.0011898
4-1	6.0606667	3.8917200	8.2296133	0.0000000
5-1	3.4887025	1.2464375	5.7309675	0.0000604
6-1	3.6490999	1.2257088	6.0724910	0.0001250
7-1	-4.7856995	-6.8833010	-2.6880981	0.0000000
8-1	0.2032467	-2.1100139	2.5165074	0.9999991
9-1	3.5006037	1.4610896	5.5401178	0.0000051
3-2	1.2906478	-0.8819595	3.4632551	0.6469505
4-2	4.2782342	2.2384331	6.3180354	0.0000000
5-2	1.7062701	-0.4113256	3.8238657	0.2289151
6-2	1.8666675	-0.4418566	4.1751915	0.2247179
7-2	-6.5681320	-8.5319016	-4.6043624	0.0000000
8-2	-1.5791857	-3.7718174	0.6134461	0.3780824
9-2	1.7181713	-0.1834272	3.6197698	0.1133698
4-3	2.9875864	0.5929682	5.3822046	0.0036891
5-3	0.4156222	-2.0456008	2.8768453	0.9998509
6-3	0.5760196	-2.0512777	3.2033169	0.9989678
7-3	-7.8587798	-10.1889726	-5.5285870	0.0000000
8-3	-2.8698335	-5.3959059	-0.3437612	0.0129870
9-3	0.4275235	-1.8505205	2.7055674	0.9996709
5-4	-2.5719642	-4.9167851	-0.2271433	0.0195776
6-4	-2.4115668	-4.9301492	0.1070156	0.0727725
7-4	-10.8463662	-13.0532565	-8.6394759	0.0000000
8-4	-5.8574199	-8.2702207	-3.4446191	0.0000000
9-4	-2.5600629	-4.7118182	-0.4083077	0.0072053
6-5	0.1603974	-2.4215941	2.7423888	0.9999999
7-5	-8.2744020	-10.5533903	-5.9954138	0.0000000
8-5	-3.2854558	-5.7643729	-0.8065386	0.0014230
9-5	0.0119012	-2.2137388	2.2375412	1.0000000
7-6	-8.4347994	-10.8922085	-5.9773903	0.0000000
8-6	-3.4458531	-6.0897333	-0.8019729	0.0018753
9-6	-0.1484962	-2.5565131	2.2595207	0.9999999
8-7	4.9889463	2.6400721	7.3378205	0.0000000
9-7	8.2863033	6.2064829	10.3661236	0.0000000
9-8	3.2973570	1.0002074	5.5945065	0.0003344

NULL

Week 8

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3885.48	8	485.68	48.96	< .001
Residuals	4117.15	415	9.92
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 415) = 48.96, p < .001; Eta2 = 0.49, 95% CI [0.43, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.2807329	-0.4678349	3.0293007	0.3542639
3-1	3.1234714	0.9997833	5.2471594	0.0002052
4-1	5.7128110	3.7385833	7.6870386	0.0000000
5-1	3.0594896	1.0185258	5.1004533	0.0001377
6-1	4.6744480	2.4686189	6.8802771	0.0000000
7-1	-5.2376744	-7.1469619	-3.3283869	0.0000000
8-1	0.2075570	-1.8980287	2.3131428	0.9999977
9-1	2.7042240	0.8478090	4.5606390	0.0002497
3-2	1.8427385	-0.1721370	3.8576139	0.1038711
4-2	4.4320781	2.5754018	6.2887544	0.0000000
5-2	1.7787567	-0.1487301	3.7062434	0.0971730
6-2	3.3937151	1.2924407	5.4949895	0.0000249
7-2	-6.5184073	-8.3058778	-4.7309368	0.0000000
8-2	-1.0731759	-3.0689623	0.9226106	0.7605111
9-2	1.4234911	-0.3073898	3.1543721	0.2050231
4-3	2.5893396	0.3757885	4.8028907	0.0090071
5-3	-0.0639818	-2.3372540	2.2092904	1.0000000
6-3	1.5509766	-0.8714015	3.9733547	0.5466755
7-3	-8.3611458	-10.5169779	-6.2053136	0.0000000
8-3	-2.9159143	-5.2473784	-0.5844502	0.0035586
9-3	-0.4192473	-2.5283965	1.6899019	0.9994959
5-4	-2.6533214	-4.7876341	-0.5190087	0.0038898
6-4	-1.0383630	-3.3308376	1.2541116	0.8927677
7-4	-10.9504854	-12.9592502	-8.9417205	0.0000000
8-4	-5.5052539	-7.7014436	-3.3090643	0.0000000
9-4	-3.0085869	-4.9671666	-1.0500073	0.0000808
6-5	1.6149584	-0.7352326	3.9651494	0.4456218
7-5	-8.2971640	-10.3715541	-6.2227738	0.0000000
8-5	-2.8519326	-5.1083029	-0.5955622	0.0030363
9-5	-0.3552655	-2.3810968	1.6705658	0.9998018
7-6	-9.9121224	-12.1489155	-7.6753292	0.0000000
8-6	-4.4668909	-6.8734146	-2.0603673	0.0000005
9-6	-1.9702239	-4.1620591	0.2216112	0.1175650
8-7	5.4452314	3.3072294	7.5832335	0.0000000
9-7	7.9418984	6.0487957	9.8350012	0.0000000
9-8	2.4966670	0.4057460	4.5875880	0.0068655

NULL

Week 9

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3444.00	7	492.00	43.32	< .001
Residuals	4292.79	378	11.36
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(7, 378) = 43.32, p < .001; Eta2 = 0.45, 95% CI [0.38, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.7840772	-0.0445944	3.6127489	0.0616839
4-1	5.8825397	3.8178705	7.9472089	0.0000000
5-1	3.4552961	1.3208335	5.5897586	0.0000330
6-1	4.3563931	2.0250534	6.6877329	0.0000007
7-1	-4.6822250	-6.7000122	-2.6644378	0.0000000
8-1	0.3625663	-1.8394787	2.5646112	0.9996492
9-1	3.0562560	1.1147966	4.9977154	0.0000627
4-2	4.0984625	2.1567298	6.0401951	0.0000000
5-2	1.6712189	-0.3445682	3.6870059	0.1873015
6-2	2.5723159	0.3491173	4.7955145	0.0110422
7-2	-6.4663022	-8.3581088	-4.5744956	0.0000000
8-2	-1.4215109	-3.5087265	0.6657046	0.4325370
9-2	1.2721788	-0.5379957	3.0823533	0.3898328
5-4	-2.4272436	-4.6593315	-0.1951557	0.0222853
6-4	-1.5261465	-3.9471858	0.8948927	0.5370390
7-4	-10.5647646	-12.6855548	-8.4439744	0.0000000
8-4	-5.5199734	-7.8167729	-3.2231739	0.0000000
9-4	-2.8262837	-4.8745880	-0.7779793	0.0008462
6-5	0.9010971	-1.5797299	3.3819240	0.9550715
7-5	-8.1375210	-10.3263158	-5.9487262	0.0000000
8-5	-3.0927298	-5.4524669	-0.7329927	0.0019751
9-5	-0.3990400	-2.5176769	1.7195968	0.9991503
7-6	-9.0386181	-11.4198021	-6.6574341	0.0000000
8-6	-3.9938269	-6.5330342	-1.4546195	0.0000639
9-6	-1.3001371	-3.6169964	1.0167221	0.6805027
8-7	5.0447912	2.7900420	7.2995405	0.0000000
9-7	7.7384810	5.7374420	9.7395199	0.0000000
9-8	2.6936897	0.5069813	4.8803982	0.0049131

NULL

Week 10

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3377.74	6	562.96	49.48	< .001
Residuals	3629.66	319	11.38
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 319) = 49.48, p < .001; Eta2 = 0.48, 95% CI [0.42, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.6025897	-0.1793428	3.3845222	0.1097490
4-1	5.6238990	3.6120008	7.6357972	0.0000000
5-1	3.4843170	1.4044093	5.5642248	0.0000224
6-1	4.1188292	1.8470763	6.3905822	0.0000030
7-1	-5.1914413	-7.1912756	-3.1916069	0.0000000
8-1	-0.1168303	-2.2625931	2.0289324	0.9999985
4-2	4.0213093	2.1292055	5.9134132	0.0000000
5-2	1.8817274	-0.0825381	3.8459928	0.0703603
6-2	2.5162396	0.3498638	4.6826153	0.0113757
7-2	-6.7940309	-8.6733021	-4.9147598	0.0000000
8-2	-1.7194200	-3.7532884	0.3144483	0.1596780
5-4	-2.1395819	-4.3146198	0.0354559	0.0572127
6-4	-1.5050697	-3.8642295	0.8540900	0.4864515
7-4	-10.8153402	-12.9139376	-8.7167429	0.0000000
8-4	-5.7407293	-7.9788248	-3.5026339	0.0000000
6-5	0.6345122	-1.7829072	3.0519316	0.9868434
7-5	-8.6757583	-10.8396420	-6.5118746	0.0000000
8-5	-3.6011474	-5.9005719	-1.3017229	0.0001001
7-6	-9.3102705	-11.6591507	-6.9613904	0.0000000
8-6	-4.2356596	-6.7099672	-1.7613519	0.0000133
8-7	5.0746109	2.8473538	7.3018681	0.0000000

NULL

Week 11

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2693.67	5	538.73	44.15	< .001
Residuals	3502.29	287	12.20
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 287) = 44.15, p < .001; Eta2 = 0.43, 95% CI [0.36, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.1488052	-0.6355239	2.9331343	0.4371827
4-1	4.9752496	2.9606455	6.9898537	0.0000000
5-1	2.6961552	0.6134501	4.7788603	0.0033161
7-1	-5.4980151	-7.5382247	-3.4578056	0.0000000
8-1	-0.0547278	-2.2033765	2.0939209	0.9999997
4-2	3.8264444	1.9317958	5.7210931	0.0000003
5-2	1.5473500	-0.4195573	3.5142573	0.2153823
7-2	-6.6468203	-8.5686733	-4.7249674	0.0000000
8-2	-1.2035330	-3.2401368	0.8330708	0.5358900
5-4	-2.2790944	-4.4570576	-0.1011313	0.0342957
7-4	-10.4732648	-12.6106272	-8.3359024	0.0000000
8-4	-5.0299775	-7.2710830	-2.7888719	0.0000000
7-5	-8.1941703	-10.3958399	-5.9925008	0.0000000
8-5	-2.7508830	-5.0534001	-0.4483659	0.0090365
8-7	5.4432873	3.1791363	7.7074383	0.0000000

NULL

Week 12

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2391.68	5	478.34	37.73	< .001
Residuals	3587.92	283	12.68
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 283) = 37.73, p < .001; Eta2 = 0.40, 95% CI [0.32, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.2131386	-0.6057666	3.0320439	0.3958593
4-1	4.6983579	2.6447155	6.7520003	0.0000000
5-1	2.7075118	0.5844487	4.8305748	0.0040528
7-1	-5.4298120	-7.5685752	-3.2910488	0.0000000
8-1	0.6123905	-1.5778940	2.8026750	0.9669437
4-2	3.4852192	1.5538567	5.4165817	0.0000063
5-2	1.4943731	-0.5106482	3.4993944	0.2705739
7-2	-6.6429506	-8.6645891	-4.6213122	0.0000000
8-2	-0.6007481	-2.6768165	1.4753203	0.9617013
5-4	-1.9908461	-4.2110131	0.2293209	0.1073938
7-4	-10.1281699	-12.3633550	-7.8929847	0.0000000
8-4	-4.0859673	-6.3705003	-1.8014344	0.0000079
7-5	-8.1373238	-10.4364545	-5.8381930	0.0000000
8-5	-2.0951212	-4.4422557	0.2520133	0.1104475
8-7	6.0422025	3.6808572	8.4035479	0.0000000

NULL

Week 13

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2136.90	5	427.38	34.48	< .001
Residuals	3482.75	281	12.39
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 281) = 34.48, p < .001; Eta2 = 0.38, 95% CI [0.30, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.8202253	0.0217241	3.6187266	0.0454187
4-1	5.0626638	3.0320586	7.0932689	0.0000000
5-1	2.9357366	0.8364895	5.0349836	0.0010767
7-1	-4.5808870	-6.7288401	-2.4329339	0.0000000
8-1	1.0448255	-1.1208889	3.2105399	0.7364483
4-2	3.2424384	1.3327415	5.1521354	0.0000271
5-2	1.1155112	-0.8670183	3.0980407	0.5895927
7-2	-6.4011123	-8.4351447	-4.3670800	0.0000000
8-2	-0.7753998	-2.8281794	1.2773798	0.8875375
5-4	-2.1269272	-4.3221889	0.0683344	0.0636439
7-4	-9.6435508	-11.8854335	-7.4016680	0.0000000
8-4	-4.0178382	-6.2767439	-1.7589326	0.0000091
7-5	-7.5166236	-9.8208629	-5.2123842	0.0000000
8-5	-1.8909110	-4.2117159	0.4298939	0.1824823
8-7	5.6257125	3.2607601	7.9906649	0.0000000

NULL

Week 14

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2274.11	5	454.82	35.00	< .001
Residuals	3652.07	281	13.00
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 281) = 35.00, p < .001; Eta2 = 0.38, 95% CI [0.31, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.8145349	-0.0271657	3.6562356	0.0561208
4-1	5.0294719	2.9500922	7.1088515	0.0000000
5-1	3.4078177	1.2581474	5.5574880	0.0001165
7-1	-4.7488311	-6.9483773	-2.5492848	0.0000000
8-1	1.6484716	-0.5692626	3.8662058	0.2734200
4-2	3.2149369	1.2593697	5.1705042	0.0000551
5-2	1.5932828	-0.4368664	3.6234320	0.2176493
7-2	-6.5633660	-8.6462552	-4.4804769	0.0000000
8-2	-0.1660634	-2.2681500	1.9360233	0.9999175
5-4	-1.6216542	-3.8696453	0.6263370	0.3061123
7-4	-9.7783030	-12.0740350	-7.4825709	0.0000000
8-4	-3.3810003	-5.6941641	-1.0678365	0.0005219
7-5	-8.1566488	-10.5162352	-5.7970623	0.0000000
8-5	-1.7593461	-4.1358960	0.6172038	0.2777916
8-7	6.3973027	3.9755449	8.8190605	0.0000000

NULL

Week 15

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	717.01	4	179.25	14.44	< .001
Residuals	3042.00	245	12.42
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(4, 245) = 14.44, p < .001; Eta2 = 0.19, 95% CI [0.11, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.7484598	0.0244375	3.4724821	0.0450075
4-1	4.9859913	3.0394769	6.9325057	0.0000000
5-1	3.3947897	1.3824760	5.4071034	0.0000564
8-1	1.2848383	-0.7911903	3.3608668	0.4353283
4-2	3.2375315	1.4069183	5.0681447	0.0000206
5-2	1.6463299	-0.2540998	3.5467595	0.1240693
8-2	-0.4636215	-2.4313921	1.5041490	0.9669833
5-4	-1.5912017	-3.6955539	0.5131506	0.2329356
8-4	-3.7011531	-5.8665136	-1.5357925	0.0000430
8-5	-2.1099514	-4.3346479	0.1147451	0.0722907

NULL

Week 16

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	659.62	4	164.91	12.57	< .001
Residuals	3213.66	245	13.12
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(4, 245) = 12.57, p < .001; Eta2 = 0.17, 95% CI [0.10, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.6941179	-0.0778800	3.4661159	0.0685514
4-1	4.8815354	2.8808538	6.8822169	0.0000000
5-1	2.6914234	0.6231115	4.7597353	0.0038159
8-1	0.9371752	-1.1966246	3.0709750	0.7473321
4-2	3.1874174	1.3058624	5.0689725	0.0000518
5-2	0.9973055	-0.9560088	2.9506198	0.6262262
8-2	-0.7569428	-2.7794719	1.2655864	0.8419806
5-4	-2.1901119	-4.3530236	-0.0272003	0.0455485
8-4	-3.9443602	-6.1699779	-1.7187425	0.0000196
8-5	-1.7542482	-4.0408530	0.5323566	0.2197905

NULL

Week 17

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	610.42	4	152.60	12.10	< .001
Residuals	3089.51	245	12.61
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(4, 245) = 12.10, p < .001; Eta2 = 0.16, 95% CI [0.09, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.4325892	-0.3048441	3.1700225	0.1595922
4-1	4.2956943	2.3340382	6.2573505	0.0000001
5-1	3.0850180	1.0570506	5.1129853	0.0003874
8-1	0.3393979	-1.7527800	2.4315757	0.9917826
4-2	2.8631051	1.0182517	4.7079585	0.0002746
5-2	1.6524287	-0.2627842	3.5676416	0.1267937
8-2	-1.0931914	-3.0762691	0.8898863	0.5537376
5-4	-1.2106764	-3.3313982	0.9100454	0.5188128
8-4	-3.9562965	-6.1385012	-1.7740918	0.0000117
8-5	-2.7456201	-4.9876223	-0.5036179	0.0078322

NULL

Week 18

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	542.41	4	135.60	9.59	< .001
Residuals	3462.98	245	14.13
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 9.59, p < .001; Eta2 = 0.14, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.2857247	-0.5537254	3.1251749	0.3089797
4-1	4.2072503	2.1304116	6.2840890	0.0000007
5-1	2.4461083	0.2990649	4.5931518	0.0165832
8-1	0.3097095	-1.9053146	2.5247337	0.9953551
4-2	2.9215255	0.9683479	4.8747031	0.0005131
5-2	1.1603836	-0.8672849	3.1880520	0.5163397
8-2	-0.9760152	-3.0755332	1.1235028	0.7052571
5-4	-1.7611419	-4.0063861	0.4841022	0.2003107
8-4	-3.8975407	-6.2078779	-1.5872035	0.0000564
8-5	-2.1363988	-4.5100446	0.2372471	0.1000765

NULL

Week 19

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	352.50	3	117.50	7.01	< .001
Residuals	3367.13	201	16.75
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(3, 201) = 7.01, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.7530965	-0.1346296	3.6408226	0.0792029
5-1	2.0594694	-0.1439228	4.2628615	0.0763012
8-1	-1.4776344	-3.7507914	0.7955226	0.3348049
5-2	0.3063728	-1.7745113	2.3872570	0.9810653
8-2	-3.2307309	-5.3853503	-1.0761115	0.0007969
8-5	-3.5371038	-5.9730454	-1.1011621	0.0012579

NULL

Week 20

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	342.47	3	114.16	7.74	< .001
Residuals	2965.44	201	14.75
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(3, 201) = 7.74, p < .001; Eta2 = 0.10, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.5693202	0.7977711	4.3408692	0.0012786
5-1	2.5169071	0.4491192	4.5846950	0.0099821
8-1	-0.1238442	-2.2571034	2.0094150	0.9987841
5-2	-0.0524130	-2.0052325	1.9004064	0.9998790
8-2	-2.6931643	-4.7151812	-0.6711475	0.0037793
8-5	-2.6407513	-4.9267768	-0.3547258	0.0163157

NULL

Week 21

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	205.78	3	68.59	5.29	0.002
Residuals	2605.91	201	12.96
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(3, 201) = 5.29, p = 0.002; Eta2 = 0.07, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.7139256	0.0532347	3.3746165	0.0401987
5-1	1.1563063	-0.7820857	3.0946984	0.4124364
8-1	-0.9200465	-2.9198129	1.0797199	0.6326743
5-2	-0.5576193	-2.3882373	1.2729988	0.8593337
8-2	-2.6339721	-4.5294573	-0.7384869	0.0022516
8-5	-2.0763528	-4.2193259	0.0666202	0.0613681

NULL

Week 22

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	122.44	2	61.22	4.71	0.010
Residuals	2093.88	161	13.01
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and small (F(2, 161) = 4.71, p = 0.010; Eta2 = 0.06, 95% CI [7.72e-03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.5077599	-0.0110384	3.0265582	0.0521549
8-1	-0.4923205	-2.3369465	1.3523055	0.8030924
8-2	-2.0000804	-3.7501923	-0.2499685	0.0206518

NULL

Week 23

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	247.74	2	123.87	9.03	< .001
Residuals	2180.46	159	13.71
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.03, p < .001; Eta2 = 0.10, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6592439	1.099462	4.2190261	0.0002500
8-1	0.5573332	-1.371832	2.4864980	0.7734516
8-2	-2.1019107	-3.935852	-0.2679697	0.0202167

NULL

Week 24

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	348.41	2	174.20	13.59	< .001
Residuals	2037.55	159	12.81
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 159) = 13.59, p < .001; Eta2 = 0.15, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	3.251735	1.7439341	4.7595366	0.0000028
8-1	1.114875	-0.7499989	2.9797488	0.3359255
8-2	-2.136860	-3.9096838	-0.3640369	0.0135861

NULL

Week 25

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	187.60	2	93.80	8.42	< .001
Residuals	1771.28	159	11.14
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 8.42, p < .001; Eta2 = 0.10, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.0431176	0.6372856	3.4489495	0.0021469
8-1	-0.2865323	-2.0252889	1.4522243	0.9196776
8-2	-2.3296499	-3.9825813	-0.6767185	0.0030385

NULL

Week 26

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	239.96	2	119.98	10.34	< .001
Residuals	1844.93	159	11.60
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 10.34, p < .001; Eta2 = 0.12, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6709540	1.2361945	4.1057136	0.0000575
8-1	0.7796968	-0.9948379	2.5542315	0.5531067
8-2	-1.8912572	-3.5782007	-0.2043138	0.0237860

NULL

Week 27

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	194.67	2	97.33	9.92	< .001
Residuals	1560.76	159	9.82
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.92, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.2693338	0.9496875	3.588980	0.0002185
8-1	0.1950148	-1.4371459	1.827175	0.9569196
8-2	-2.0743190	-3.6259161	-0.522722	0.0052900

NULL

Week 28

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	252.26	2	126.13	11.37	< .001
Residuals	1763.75	159	11.09
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 11.37, p < .001; Eta2 = 0.13, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.3849702	0.9821283	3.787812	0.0002610
8-1	-0.2962779	-2.0313362	1.438780	0.9140215
8-2	-2.6812481	-4.3306637	-1.031832	0.0005067

NULL

Week 29

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	191.31	2	95.65	9.18	< .001
Residuals	1655.88	159	10.41
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.18, p < .001; Eta2 = 0.10, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.4093057	1.0500389	3.7685726	0.0001341
8-1	0.8246694	-0.8564946	2.5058335	0.4785242
8-2	-1.5846363	-3.1828179	0.0135453	0.0525199

NULL

Week 30

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	267.02	2	133.51	12.26	< .001
Residuals	1731.82	159	10.89
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 12.26, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6988171	1.308733	4.0889014	0.0000262
8-1	0.3580401	-1.361239	2.0773197	0.8748953
8-2	-2.3407771	-3.975193	-0.7063613	0.0025404

NULL

Week 31

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	194.14	2	97.07	10.23	< .001
Residuals	1509.40	159	9.49
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 10.23, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.2181551	0.9204022	3.5159080	0.0002403
8-1	0.0555941	-1.5494884	1.6606767	0.9963048
8-2	-2.1625610	-3.6884166	-0.6367055	0.0028569

NULL

Week 32

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	179.33	2	89.66	8.19	< .001
Residuals	1741.05	159	10.95
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 8.19, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.0800183	0.6862368	3.4737997	0.0015653
8-1	-0.0833907	-1.8072429	1.6404616	0.9928050
8-2	-2.1634089	-3.8021717	-0.5246462	0.0059919

NULL

Week 33

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	228.91	2	114.45	12.17	< .001
Residuals	1486.24	158	9.41
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 12.17, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.6497193	1.3540466	3.9453920	0.0000092
8-1	0.9361127	-0.6617377	2.5339632	0.3505878
8-2	-1.7136065	-3.2357922	-0.1914208	0.0230872

NULL

Week 34

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	168.86	2	84.43	8.10	< .001
Residuals	1647.79	158	10.43
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 8.10, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.0817196	0.7174456	3.4459936	0.0011881
8-1	0.0668609	-1.6155901	1.7493118	0.9951379
8-2	-2.0148587	-3.6176388	-0.4120787	0.0094691

NULL

Week 35

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	194.12	2	97.06	8.94	< .001
Residuals	1714.87	158	10.85
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 8.94, p < .001; Eta2 = 0.10, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	2.1148860	0.7231223	3.5066496	0.0012520
8-1	-0.2261492	-1.9425009	1.4902026	0.9478557
8-2	-2.3410351	-3.9761106	-0.7059596	0.0025505

NULL

Week 36

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	118.82	2	59.41	4.64	0.011
Residuals	2008.24	157	12.79
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and small (F(2, 157) = 4.64, p = 0.011; Eta2 = 0.06, 95% CI [7.57e-03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	1.6843299	0.1653771	3.2032828	0.0257321
8-1	-0.1156115	-1.9854604	1.7542374	0.9882707
8-2	-1.7999414	-3.5750923	-0.0247905	0.0460723

NULL

Week 37

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.49	1	1.49	0.14	0.706
Residuals	873.80	84	10.40
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and very small (F(1, 84) = 0.14, p = 0.706; Eta2 = 1.70e-03, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.270507	-1.692742	1.151728	0.7062134

NULL

Week 38

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.65	1	1.65	0.15	0.697
Residuals	898.27	83	10.82
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and very small (F(1, 83) = 0.15, p = 0.697; Eta2 = 1.83e-03, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.2857376	-1.742008	1.170532	0.6973458

NULL

Week 39

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	10.00	1	10.00	1.08	0.303
Residuals	752.45	81	9.29
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(1, 81) = 1.08, p = 0.303; Eta2 = 0.01, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.7131589	-2.080764	0.6544464	0.3025658

NULL

Week 40

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.04	1	0.04	3.60e-03	0.952
Residuals	827.16	80	10.34
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and very small (F(1, 80) = 3.60e-03, p = 0.952; Eta2 = 4.50e-05, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.0436836	-1.404966	1.492333	0.9522975

NULL

Week 41

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	4.02	1	4.02	0.36	0.551
Residuals	705.28	63	11.19
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and very small (F(1, 63) = 0.36, p = 0.551; Eta2 = 5.67e-03, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.4975069	-2.156353	1.16134	0.5511055

NULL

Week 42

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	5.98	1	5.98	0.69	0.411
Residuals	522.96	60	8.72
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(1, 60) = 0.69, p = 0.411; Eta2 = 0.01, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.6212041	-2.121977	0.8795686	0.4109693

NULL

Week 43

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	10.93	1	10.93	1.14	0.289
Residuals	534.91	56	9.55
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(1, 56) = 1.14, p = 0.289; Eta2 = 0.02, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	-0.8701893	-2.499976	0.7595969	0.2893938

NULL

Model fitting for FE as output variable

Week 1

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	122.95	6	20.49	25.26	< .001
Residuals	225.56	278	0.81
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(6, 278) = 25.26, p < .001; Eta2 = 0.35, 95% CI [0.27, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.5688962	-1.0451906	-0.0926018	0.0082050
3-1	1.5846680	1.0061941	2.1631418	0.0000000
5-1	-0.0084235	-0.5643640	0.5475169	1.0000000
6-1	0.0671543	-0.5400647	0.6743732	0.9998980
7-1	0.0604691	-1.0897637	1.2107020	0.9999988
9-1	0.6255523	0.0877901	1.1633144	0.0112316
3-2	2.1535642	1.6047299	2.7023984	0.0000000
5-2	0.5604727	0.0354423	1.0855030	0.0278342
6-2	0.6360505	0.0569979	1.2151031	0.0209622
7-2	0.6293653	-0.5062502	1.7649809	0.6524975
9-2	1.1944485	0.6887063	1.7001907	0.0000000
5-3	-1.5930915	-2.2123107	-0.9738722	0.0000000
6-3	-1.5175137	-2.1831544	-0.8518730	0.0000000
7-3	-1.5241988	-2.7063148	-0.3420828	0.0029888
9-3	-0.9591157	-1.5620674	-0.3561639	0.0000746
6-5	0.0755778	-0.5705765	0.7217321	0.9998584
7-5	0.0688927	-1.1023614	1.2401468	0.9999976
9-5	0.6339758	0.0526079	1.2153437	0.0225658
7-6	-0.0066851	-1.2031304	1.1897601	1.0000000
9-6	0.5583980	-0.0721840	1.1889800	0.1210010
9-7	0.5650831	-0.5976526	1.7278189	0.7776320

NULL

Week 2

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	37.78	5	7.56	17.07	< .001
Residuals	124.82	282	0.44
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(5, 282) = 17.07, p < .001; Eta2 = 0.23, 95% CI [0.16, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0783269	-0.4181943	0.2615404	0.9859297
3-1	0.9866446	0.5738656	1.3994237	0.0000000
5-1	0.0851852	-0.3115149	0.4818852	0.9898017
6-1	0.1257901	-0.3029546	0.5545348	0.9593808
9-1	0.5836536	0.2161985	0.9511088	0.0001119
3-2	1.0649716	0.6733423	1.4566008	0.0000000
5-2	0.1635121	-0.2111315	0.5381557	0.8104095
6-2	0.2041170	-0.2043055	0.6125396	0.7062084
9-2	0.6619806	0.3184544	1.0055068	0.0000011
5-3	-0.9014595	-1.3433131	-0.4596059	0.0000002
6-3	-0.8608546	-1.3316897	-0.3900194	0.0000045
9-3	-0.4029910	-0.8187878	0.0128058	0.0634853
6-5	0.0406049	-0.4161993	0.4974091	0.9998524
9-5	0.4984685	0.0986293	0.8983076	0.0054231
9-6	0.4578636	0.0262127	0.8895144	0.0304208

NULL

Week 3

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	15.11	7	2.16	13.16	< .001
Residuals	54.27	331	0.16
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(7, 331) = 13.16, p < .001; Eta2 = 0.22, 95% CI [0.14, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0190678	-0.2389671	0.2008315	0.9999955
3-1	0.5489440	0.2863179	0.8115701	0.0000000
4-1	-0.0922294	-0.3405076	0.1560488	0.9491103
5-1	0.0841077	-0.1725632	0.3407786	0.9742932
6-1	0.1363428	-0.1410615	0.4137471	0.8075459
8-1	0.3146542	-0.5744799	1.2037883	0.9607119
9-1	0.3880315	0.1524941	0.6235688	0.0000225
3-2	0.5680118	0.3193146	0.8167090	0.0000000
4-2	-0.0731616	-0.3066566	0.1603334	0.9800617
5-2	0.1031756	-0.1392245	0.3455756	0.8990764
6-2	0.1554107	-0.1088449	0.4196662	0.6246359
8-2	0.3337220	-0.5513979	1.2188420	0.9449932
9-2	0.4070993	0.1872000	0.6269986	0.0000010
4-3	-0.6411734	-0.9152841	-0.3670627	0.0000000
5-3	-0.4648362	-0.7465712	-0.1831013	0.0000217
6-3	-0.4126011	-0.7133465	-0.1118558	0.0009478
8-3	-0.2342897	-1.1309804	0.6624009	0.9932005
9-3	-0.1609125	-0.4235386	0.1017136	0.5732357
5-4	0.1763371	-0.0920733	0.4447476	0.4806941
6-4	0.2285722	-0.0597286	0.5168731	0.2358513
8-4	0.4068836	-0.4857101	1.2994774	0.8612095
9-4	0.4802609	0.2319827	0.7285391	0.0000002
6-5	0.0522351	-0.2433241	0.3477943	0.9994355
8-5	0.2305465	-0.6644181	1.1255110	0.9937647
9-5	0.3039237	0.0472528	0.5605946	0.0083336
8-6	0.1783114	-0.7228183	1.0794411	0.9988199
9-6	0.2516886	-0.0257157	0.5290930	0.1068621
9-8	0.0733772	-0.8157568	0.9625113	0.9999968

NULL

Week 4

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	10.28	7	1.47	7.69	< .001
Residuals	66.27	347	0.19
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(7, 347) = 7.69, p < .001; Eta2 = 0.13, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0643986	-0.3016641	0.1728669	0.9914573
3-1	0.4329770	0.1472672	0.7186868	0.0001446
4-1	-0.1097013	-0.3775869	0.1581843	0.9165018
5-1	0.0348237	-0.2421174	0.3117649	0.9999423
6-1	0.1920941	-0.1072179	0.4914060	0.5125531
8-1	0.2688451	-0.0730185	0.6107087	0.2454916
9-1	0.2225824	-0.0339425	0.4791074	0.1433146
3-2	0.4973756	0.2265647	0.7681866	0.0000012
4-2	-0.0453027	-0.2972376	0.2066322	0.9993694
5-2	0.0992223	-0.1623210	0.3607656	0.9433100
6-2	0.2564927	-0.0286321	0.5416175	0.1133687
8-2	0.3332437	0.0037301	0.6627573	0.0452566
9-2	0.2869810	0.0471612	0.5268008	0.0072668
4-3	-0.5426783	-0.8406823	-0.2446744	0.0000015
5-3	-0.3981533	-0.7043233	-0.0919834	0.0022442
6-3	-0.2408829	-0.5674274	0.0856615	0.3249909
8-3	-0.1641319	-0.5300750	0.2018111	0.8712628
9-3	-0.2103946	-0.4982291	0.0774399	0.3369385
5-4	0.1445250	-0.1450828	0.4341328	0.7953147
6-4	0.3017954	-0.0092736	0.6128644	0.0645532
8-4	0.3785464	0.0263434	0.7307494	0.0252026
9-4	0.3322837	0.0621332	0.6024343	0.0050381
6-5	0.1572704	-0.1616302	0.4761710	0.8050801
8-5	0.2340214	-0.1251174	0.5931601	0.4920868
9-5	0.1877587	-0.0913739	0.4668913	0.4489403
8-6	0.0767510	-0.2999079	0.4534099	0.9985778
9-6	0.0304883	-0.2708524	0.3318291	0.9999869
9-8	-0.0462627	-0.3899039	0.2973786	0.9999084

NULL

Week 5

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	12.61	8	1.58	9.00	< .001
Residuals	65.21	372	0.18
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(8, 372) = 9.00, p < .001; Eta2 = 0.16, 95% CI [0.10, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0746854	-0.3072517	0.1578809	0.9855717
3-1	0.5156827	0.2379283	0.7934370	0.0000005
4-1	-0.0829098	-0.3454897	0.1796701	0.9870670
5-1	0.0339595	-0.2374966	0.3054156	0.9999849
6-1	-0.0367630	-0.3301468	0.2566209	0.9999847
7-1	0.1337154	-0.2139062	0.4813371	0.9561599
8-1	0.0432460	-0.2718499	0.3583419	0.9999691
9-1	0.2557337	0.0066285	0.5048389	0.0390968
3-2	0.5903681	0.3273450	0.8533912	0.0000000
4-2	-0.0082244	-0.2551695	0.2387208	1.0000000
5-2	0.1086449	-0.1477183	0.3650081	0.9242556
6-2	0.0379225	-0.2415552	0.3174001	0.9999717
7-2	0.2084009	-0.1275672	0.5443689	0.5895302
8-2	0.1179314	-0.1842591	0.4201220	0.9523201
9-2	0.3304191	0.0978528	0.5629854	0.0004145
4-3	-0.5985925	-0.8884930	-0.3086920	0.0000000
5-3	-0.4817232	-0.7796871	-0.1837593	0.0000250
6-3	-0.5524456	-0.8705151	-0.2343762	0.0000038
7-3	-0.3819672	-0.7506608	-0.0132737	0.0357725
8-3	-0.4724367	-0.8106371	-0.1342363	0.0005700
9-3	-0.2599490	-0.5377033	0.0178054	0.0872357
5-4	0.1168693	-0.1670026	0.4007412	0.9354094
6-4	0.0461468	-0.2587612	0.3510549	0.9999346
7-4	0.2166252	-0.1407760	0.5740265	0.6203052
8-4	0.1261558	-0.1996973	0.4520089	0.9544749
9-4	0.3386435	0.0760636	0.6012235	0.0022386
6-5	-0.0707225	-0.3833070	0.2418621	0.9986950
7-5	0.0997559	-0.2642164	0.4637283	0.9949734
8-5	0.0092865	-0.3237607	0.3423337	1.0000000
9-5	0.2217742	-0.0496819	0.4932303	0.2122810
7-6	0.1704784	-0.2101284	0.5510852	0.8983695
8-6	0.0800090	-0.2711408	0.4311587	0.9986272
9-6	0.2924967	-0.0008871	0.5858805	0.0513893
8-7	-0.0904695	-0.4880533	0.3071143	0.9986402
9-7	0.1220183	-0.2256034	0.4696400	0.9747076
9-8	0.2124877	-0.1026082	0.5275837	0.4720705

NULL

Week 6

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3.60	7	0.51	4.24	< .001
Residuals	44.41	366	0.12
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(7, 366) = 4.24, p < .001; Eta2 = 0.07, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0297135	-0.2187620	0.1593350	0.9997420
3-1	0.1760265	-0.0497545	0.4018074	0.2559558
4-1	-0.0935517	-0.3069977	0.1198943	0.8844942
5-1	0.0317821	-0.1888791	0.2524434	0.9998561
6-1	0.0220864	-0.2163995	0.2605722	0.9999929
8-1	0.0608814	-0.1687237	0.2904865	0.9926034
9-1	0.2095887	0.0088802	0.4102972	0.0335789
3-2	0.2057400	-0.0080663	0.4195462	0.0689808
4-2	-0.0638382	-0.2645750	0.1368986	0.9784025
5-2	0.0614956	-0.1468969	0.2698882	0.9859871
6-2	0.0517999	-0.1753819	0.2789816	0.9970964
8-2	0.0905949	-0.1272458	0.3084355	0.9101124
9-2	0.2393022	0.0521659	0.4264384	0.0028819
4-3	-0.2695781	-0.5052324	-0.0339238	0.0126588
5-3	-0.1442443	-0.3864532	0.0979646	0.6099023
6-3	-0.1539401	-0.4124923	0.1046121	0.6102031
8-3	-0.1151451	-0.3655293	0.1352391	0.8560854
9-3	0.0335622	-0.1906200	0.2577445	0.9998136
5-4	0.1253338	-0.1054199	0.3560876	0.7156915
6-4	0.1156380	-0.1322156	0.3634917	0.8465566
8-4	0.1544331	-0.0848876	0.3937538	0.5055836
9-4	0.3031404	0.0913862	0.5148945	0.0004385
6-5	-0.0096958	-0.2637895	0.2443979	1.0000000
8-5	0.0290993	-0.2166783	0.2748768	0.9999618
9-5	0.1778065	-0.0412186	0.3968317	0.2095874
8-6	0.0387950	-0.2231033	0.3006933	0.9998263
9-6	0.1875023	-0.0494706	0.4244752	0.2385336
9-8	0.1487073	-0.0793259	0.3767405	0.4913784

NULL

Week 7

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3.34	7	0.48	5.10	< .001
Residuals	33.83	362	0.09
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(7, 362) = 5.10, p < .001; Eta2 = 0.09, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0230717	-0.1889887	0.1428454	0.9998864
3-1	0.2152655	0.0171105	0.4134205	0.0225346
4-1	-0.0748307	-0.2621600	0.1124986	0.9263311
5-1	0.0802111	-0.1134507	0.2738728	0.9118539
6-1	-0.0084373	-0.2177426	0.2008681	1.0000000
8-1	0.1808418	-0.0206694	0.3823530	0.1152735
9-1	0.1314010	-0.0479840	0.3107859	0.3343739
3-2	0.2383372	0.0506917	0.4259827	0.0031866
4-2	-0.0517590	-0.2279342	0.1244162	0.9863390
5-2	0.1032827	-0.0796115	0.2861769	0.6731047
6-2	0.0146344	-0.1847501	0.2140188	0.9999986
8-2	0.2039135	0.0127272	0.3950998	0.0272444
9-2	0.1544726	-0.0132306	0.3221758	0.0961295
4-3	-0.2900962	-0.4969165	-0.0832759	0.0006370
5-3	-0.1350545	-0.3476273	0.0775184	0.5263417
6-3	-0.2237028	-0.4506193	0.0032137	0.0565284
8-3	-0.0344237	-0.2541716	0.1853242	0.9997475
9-3	-0.0838646	-0.2835175	0.1157884	0.9056062
5-4	0.1550417	-0.0474776	0.3575611	0.2780653
6-4	0.0663934	-0.1511335	0.2839203	0.9829334
8-4	0.2556725	0.0456344	0.4657106	0.0057897
9-4	0.2062316	0.0173185	0.3951448	0.0213967
6-5	-0.0886483	-0.3116518	0.1343551	0.9281141
8-5	0.1006308	-0.1150741	0.3163356	0.8465597
9-5	0.0511899	-0.1440043	0.2463841	0.9930791
8-6	0.1892791	-0.0405740	0.4191323	0.1942406
9-6	0.1398382	-0.0708859	0.3505623	0.4676397
9-8	-0.0494409	-0.2524253	0.1535435	0.9956080

NULL

Week 8

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3.46	7	0.49	6.41	< .001
Residuals	27.62	358	0.08
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(7, 358) = 6.41, p < .001; Eta2 = 0.11, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0401325	-0.1106437	0.1909087	0.9924211
3-1	0.3042080	0.1194474	0.4889685	0.0000223
4-1	0.0287497	-0.1414847	0.1989842	0.9995837
5-1	0.0624397	-0.1135494	0.2384287	0.9603052
6-1	-0.0331105	-0.2253323	0.1591112	0.9995257
8-1	0.1128815	-0.0686798	0.2944427	0.5548361
9-1	0.1987388	0.0349275	0.3625501	0.0060570
3-2	0.2640754	0.0886100	0.4395409	0.0001657
4-2	-0.0113828	-0.1714810	0.1487154	0.9999989
5-2	0.0223071	-0.1438969	0.1885112	0.9999105
6-2	-0.0732431	-0.2565485	0.1100624	0.9262001
8-2	0.0727489	-0.0993445	0.2448424	0.9026281
9-2	0.1586063	0.0053554	0.3118571	0.0365801
4-3	-0.2754582	-0.4679016	-0.0830149	0.0004409
5-3	-0.2417683	-0.4393203	-0.0442163	0.0053875
6-3	-0.3373185	-0.5494597	-0.1251773	0.0000506
8-3	-0.1913265	-0.3938584	0.0112054	0.0796860
9-3	-0.1054692	-0.2922547	0.0813163	0.6731496
5-4	0.0336899	-0.1503484	0.2177283	0.9992924
6-4	-0.0618603	-0.2614779	0.1377573	0.9813581
8-4	0.0841317	-0.1052422	0.2735056	0.8769108
9-4	0.1699891	-0.0024410	0.3424191	0.0565244
6-5	-0.0955502	-0.3000974	0.1089970	0.8456217
8-5	0.0504418	-0.1441214	0.2450050	0.9935438
9-5	0.1362991	-0.0418146	0.3144129	0.2785490
8-6	0.1459920	-0.0633687	0.3553527	0.4001081
9-6	0.2318493	0.0376804	0.4260183	0.0074637
9-8	0.0858573	-0.0977642	0.2694788	0.8449695

NULL

Week 9

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3.15	7	0.45	5.44	< .001
Residuals	28.29	342	0.08
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(7, 342) = 5.44, p < .001; Eta2 = 0.10, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0224002	-0.1785703	0.1337699	0.9998595
4-1	-0.0191770	-0.1955015	0.1571475	0.9999785
5-1	0.0210683	-0.1612166	0.2033532	0.9999674
6-1	-0.0368687	-0.2359671	0.1622296	0.9992349
7-1	0.2624926	0.0190701	0.5059150	0.0243649
8-1	0.1630569	-0.0249996	0.3511134	0.1439238
9-1	0.1819526	0.0131058	0.3507994	0.0245344
4-2	0.0032232	-0.1626024	0.1690488	1.0000000
5-2	0.0434685	-0.1286814	0.2156184	0.9944951
6-2	-0.0144685	-0.2043315	0.1753945	0.9999981
7-2	0.2848928	0.0489642	0.5208214	0.0064598
8-2	0.1854571	0.0072072	0.3637071	0.0347949
9-2	0.2043528	0.0465014	0.3622041	0.0024069
5-4	0.0402453	-0.1503769	0.2308675	0.9982124
6-4	-0.0176917	-0.2244505	0.1890671	0.9999959
7-4	0.2816696	0.0319427	0.5313965	0.0149399
8-4	0.1822339	-0.0139147	0.3783825	0.0902603
9-4	0.2011296	0.0233143	0.3789449	0.0144620
6-5	-0.0579370	-0.2698017	0.1539277	0.9910559
7-5	0.2414243	-0.0125462	0.4953947	0.0759152
8-5	0.1419886	-0.0595349	0.3435122	0.3858435
9-5	0.1608843	-0.0228430	0.3446116	0.1352844
7-6	0.2993613	0.0330656	0.5656569	0.0155157
8-6	0.1999256	-0.0169248	0.4167761	0.0954494
9-6	0.2188213	0.0184015	0.4192411	0.0213866
8-7	-0.0994356	-0.3575799	0.1587086	0.9386295
9-7	-0.0805400	-0.3250445	0.1639645	0.9735717
9-8	0.0188957	-0.1705593	0.2083506	0.9999881

NULL

Week 10

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.23	5	0.45	7.45	< .001
Residuals	16.38	274	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(5, 274) = 7.45, p < .001; Eta2 = 0.12, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0125851	-0.1123287	0.1374989	0.9997263
4-1	0.0826406	-0.0583939	0.2236750	0.5450757
5-1	0.0491226	-0.0966793	0.1949245	0.9279838
6-1	0.0579639	-0.1012864	0.2172142	0.9023389
8-1	0.2864402	0.1360219	0.4368586	0.0000015
4-2	0.0700554	-0.0625814	0.2026922	0.6543591
5-2	0.0365375	-0.1011578	0.1742329	0.9736225
6-2	0.0453788	-0.1064845	0.1972421	0.9560655
8-2	0.2738551	0.1312806	0.4164296	0.0000012
5-4	-0.0335179	-0.1859885	0.1189526	0.9886329
6-4	-0.0246767	-0.1900542	0.1407009	0.9981557
8-4	0.2037997	0.0469088	0.3606906	0.0031764
6-5	0.0088413	-0.1606203	0.1783028	0.9999895
8-5	0.2373176	0.0761275	0.3985077	0.0004625
8-6	0.2284763	0.0550269	0.4019257	0.0026206

NULL

Week 11

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.01	4	0.25	4.24	0.002
Residuals	14.56	245	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 4.24, p = 0.002; Eta2 = 0.06, 95% CI [0.01, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0301881	-0.0890891	0.1494654	0.9573153
4-1	0.0690910	-0.0655795	0.2037614	0.6218616
5-1	0.0792649	-0.0599580	0.2184877	0.5215465
8-1	0.2022579	0.0586270	0.3458889	0.0013068
4-2	0.0389028	-0.0877490	0.1655546	0.9164937
5-2	0.0490767	-0.0824054	0.1805588	0.8432751
8-2	0.1720698	0.0359287	0.3082109	0.0054463
5-4	0.0101739	-0.1354167	0.1557644	0.9996947
8-4	0.1331669	-0.0166445	0.2829784	0.1075595
8-5	0.1229931	-0.0309235	0.2769097	0.1846365

NULL

Week 12

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.53	4	0.13	2.11	0.080
Residuals	15.34	245	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(4, 245) = 2.11, p = 0.080; Eta2 = 0.03, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0381984	-0.0842157	0.1606125	0.9119637
4-1	0.0940212	-0.0441910	0.2322333	0.3366096
5-1	0.0971867	-0.0456976	0.2400709	0.3367443
8-1	0.1326446	-0.0147637	0.2800529	0.1002100
4-2	0.0558227	-0.0741599	0.1858053	0.7627018
5-2	0.0589883	-0.0759517	0.1939282	0.7506124
8-2	0.0944462	-0.0452753	0.2341676	0.3431487
5-4	0.0031655	-0.1462539	0.1525849	0.9999974
8-4	0.0386234	-0.1151279	0.1923748	0.9584369
8-5	0.0354579	-0.1225065	0.1934224	0.9723104

NULL

Week 13

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.81	5	0.36	5.70	< .001
Residuals	16.29	257	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(5, 257) = 5.70, p < .001; Eta2 = 0.10, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0370334	-0.0916800	0.1657468	0.9624842
4-1	0.1076629	-0.0376615	0.2529872	0.2765917
5-1	0.0688905	-0.0813463	0.2191274	0.7756455
7-1	0.3413801	0.1184199	0.5643404	0.0002318
8-1	0.1843335	0.0293398	0.3393272	0.0095627
4-2	0.0706295	-0.0660418	0.2073008	0.6749114
5-2	0.0318571	-0.1100266	0.1737409	0.9874492
7-2	0.3043467	0.0869274	0.5217660	0.0010724
8-2	0.1473001	0.0003888	0.2942114	0.0489657
5-4	-0.0387723	-0.1958806	0.1183360	0.9807967
7-4	0.2337173	0.0060702	0.4613643	0.0403831
8-4	0.0766706	-0.0849925	0.2383337	0.7497897
7-5	0.2724896	0.0416755	0.5033036	0.0103921
8-5	0.1154429	-0.0506502	0.2815360	0.3475982
8-7	-0.1570467	-0.3909848	0.0768915	0.3877038

NULL

Week 14

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.99	4	0.25	5.06	< .001
Residuals	11.98	245	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 5.06, p < .001; Eta2 = 0.08, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0294628	-0.0787197	0.1376453	0.9447243
4-1	0.0950264	-0.0271174	0.2171703	0.2074139
5-1	0.0120160	-0.1142568	0.1382888	0.9989657
8-1	0.1871020	0.0568312	0.3173729	0.0009737
4-2	0.0655636	-0.0493074	0.1804347	0.5190274
5-2	-0.0174468	-0.1366989	0.1018053	0.9944687
8-2	0.1576392	0.0341615	0.2811169	0.0048288
5-4	-0.0830104	-0.2150586	0.0490378	0.4189383
8-4	0.0920756	-0.0438009	0.2279521	0.3405878
8-5	0.1750860	0.0354862	0.3146858	0.0059642

NULL

Week 15

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.21	4	0.30	5.81	< .001
Residuals	12.80	245	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 5.81, p < .001; Eta2 = 0.09, 95% CI [0.03, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0267661	-0.0850764	0.1386086	0.9650667
4-1	0.0932446	-0.0330316	0.2195208	0.2551072
5-1	-0.0038074	-0.1343522	0.1267374	0.9999906
8-1	0.2027312	0.0680530	0.3374094	0.0004621
4-2	0.0664785	-0.0522789	0.1852359	0.5385348
5-2	-0.0305735	-0.1538601	0.0927131	0.9603078
8-2	0.1759651	0.0483099	0.3036202	0.0017743
5-4	-0.0970520	-0.2335676	0.0394636	0.2920134
8-4	0.1094866	-0.0309869	0.2499600	0.2058193
8-5	0.2065386	0.0622159	0.3508613	0.0010283

NULL

Week 16

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.89	4	0.22	4.41	0.002
Residuals	12.40	245	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 4.41, p = 0.002; Eta2 = 0.07, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0321875	-0.0778866	0.1422616	0.9293241
4-1	0.1009875	-0.0232921	0.2252671	0.1710322
5-1	0.0474709	-0.0810098	0.1759516	0.8481297
8-1	0.1843869	0.0518382	0.3169357	0.0015590
4-2	0.0688000	-0.0480797	0.1856796	0.4875017
5-2	0.0152834	-0.1060538	0.1366206	0.9969008
8-2	0.1521994	0.0265627	0.2778362	0.0088273
5-4	-0.0535165	-0.1878737	0.0808406	0.8092031
8-4	0.0833995	-0.0548528	0.2216518	0.4621021
8-5	0.1369160	-0.0051247	0.2789568	0.0648619

NULL

Week 17

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.61	4	0.40	7.53	< .001
Residuals	13.05	244	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 244) = 7.53, p < .001; Eta2 = 0.11, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0704788	-0.0426551	0.1836126	0.4284220
4-1	0.1350524	0.0073182	0.2627867	0.0323567
5-1	0.0018939	-0.1301582	0.1339461	0.9999994
8-1	0.2378244	0.1004200	0.3752289	0.0000330
4-2	0.0645737	-0.0555549	0.1847023	0.5782204
5-2	-0.0685848	-0.1932949	0.0561253	0.5560432
8-2	0.1673457	0.0369815	0.2977099	0.0045184
5-4	-0.1331585	-0.2712504	0.0049334	0.0646988
8-4	0.1027720	-0.0404467	0.2459907	0.2827725
8-5	0.2359305	0.0888478	0.3830132	0.0001512

NULL

Week 18

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.05	4	0.26	5.05	< .001
Residuals	12.77	245	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(4, 245) = 5.05, p < .001; Eta2 = 0.08, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0743556	-0.0373484	0.1860596	0.3590665
4-1	0.1185116	-0.0076082	0.2446315	0.0767841
5-1	0.0256495	-0.1047337	0.1560326	0.9829853
8-1	0.1998101	0.0652986	0.3343215	0.0005743
4-2	0.0441561	-0.0744542	0.1627664	0.8445406
5-2	-0.0487061	-0.1718400	0.0744278	0.8131128
8-2	0.1254545	-0.0020426	0.2529516	0.0562001
5-4	-0.0928622	-0.2292088	0.0434844	0.3353906
8-4	0.0812984	-0.0590011	0.2215979	0.5036312
8-5	0.1741606	0.0300166	0.3183046	0.0090839

NULL

Week 19

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	3.67	3	1.22	13.96	< .001
Residuals	17.54	200	0.09
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(3, 200) = 13.96, p < .001; Eta2 = 0.17, 95% CI [0.09, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0463792	-0.0902151	0.1829734	0.8154064
5-1	0.0592179	-0.1002176	0.2186534	0.7710052
8-1	0.3853322	0.2194344	0.5512299	0.0000000
5-2	0.0128387	-0.1377322	0.1634097	0.9961849
8-2	0.3389530	0.1815554	0.4963506	0.0000005
8-5	0.3261143	0.1485313	0.5036972	0.0000221

NULL

Week 20

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.28	3	0.43	8.02	< .001
Residuals	10.62	199	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(3, 199) = 8.02, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0107738	-0.1173317	0.0957842	0.9936859
5-1	-0.0318647	-0.1562413	0.0925119	0.9105598
8-1	0.1987025	0.0681268	0.3292782	0.0006412
5-2	-0.0210909	-0.1385522	0.0963703	0.9665471
8-2	0.2094762	0.0854697	0.3334828	0.0001139
8-5	0.2305672	0.0909514	0.3701830	0.0001703

NULL

Week 21

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.49	3	0.16	1.19	0.315
Residuals	27.82	201	0.14
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically not significant and small (F(3, 201) = 1.19, p = 0.315; Eta2 = 0.02, 95% CI [0.00, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	-0.0088091	-0.1803826	0.1627645	0.9991573
5-1	-0.0110215	-0.2112856	0.1892426	0.9989632
8-1	0.1221026	-0.0845024	0.3287076	0.4208793
5-2	-0.0022124	-0.1913419	0.1869171	0.9999900
8-2	0.1309117	-0.0649196	0.3267429	0.3099365
8-5	0.1331241	-0.0882762	0.3545244	0.4052391

NULL

Week 22

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.83	2	0.91	15.40	< .001
Residuals	9.55	161	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 161) = 15.40, p < .001; Eta2 = 0.16, 95% CI [0.08, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1018006	-0.0007470	0.2043482	0.0521598
8-1	0.2917718	0.1672246	0.4163189	0.0000004
8-2	0.1899711	0.0718055	0.3081367	0.0005908

NULL

Week 23

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.51	2	0.76	11.49	< .001
Residuals	10.46	159	0.07
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 11.49, p < .001; Eta2 = 0.13, 95% CI [0.05, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0603122	-0.0477256	0.1683500	0.3857652
8-1	0.2652194	0.1315964	0.3988423	0.0000169
8-2	0.2049072	0.0778799	0.3319345	0.0005650

NULL

Week 24

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.13	2	0.57	9.38	< .001
Residuals	9.59	159	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.38, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0118620	-0.0915551	0.1152792	0.9602294
8-1	0.2138422	0.0859342	0.3417503	0.0003366
8-2	0.2019802	0.0803857	0.3235747	0.0003704

NULL

Week 25

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.24	2	0.62	12.35	< .001
Residuals	7.99	159	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 12.35, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0774245	-0.0170154	0.1718643	0.1310160
8-1	0.2445609	0.1277561	0.3613657	0.0000055
8-2	0.1671364	0.0560971	0.2781758	0.0014087

NULL

Week 26

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.25	2	0.63	9.81	< .001
Residuals	10.15	159	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 9.81, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0334467	-0.0729898	0.1398831	0.7380133
8-1	0.2344008	0.1027585	0.3660432	0.0001244
8-2	0.2009542	0.0758097	0.3260986	0.0006021

NULL

Week 27

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.23	2	0.62	13.72	< .001
Residuals	7.14	159	0.04
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 159) = 13.72, p < .001; Eta2 = 0.15, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0791902	-0.0100476	0.1684280	0.0931648
8-1	0.2438177	0.1334469	0.3541885	0.0000016
8-2	0.1646275	0.0597046	0.2695504	0.0008254

NULL

Week 28

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.35	2	0.67	12.39	< .001
Residuals	8.66	159	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 12.39, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0938740	-0.0044165	0.1921644	0.0646129
8-1	0.2557312	0.1341639	0.3772984	0.0000050
8-2	0.1618572	0.0462905	0.2774239	0.0032555

NULL

Week 29

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.21	2	0.60	13.26	< .001
Residuals	7.25	159	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 159) = 13.26, p < .001; Eta2 = 0.14, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1168061	0.0268933	0.2067188	0.0069946
8-1	0.2390735	0.1278679	0.3502791	0.0000030
8-2	0.1222675	0.0165510	0.2279840	0.0188720

NULL

Week 30

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.24	2	0.62	12.05	< .001
Residuals	8.18	159	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 12.05, p < .001; Eta2 = 0.13, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0910687	-0.0044933	0.1866308	0.0653701
8-1	0.2451968	0.1270040	0.3633896	0.0000067
8-2	0.1541281	0.0417693	0.2664868	0.0040645

NULL

Week 31

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.18	2	0.59	10.23	< .001
Residuals	9.18	159	0.06
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 10.23, p < .001; Eta2 = 0.11, 95% CI [0.04, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0972197	-0.0039996	0.1984391	0.0626807
8-1	0.2390992	0.1139095	0.3642890	0.0000359
8-2	0.1418795	0.0228691	0.2608899	0.0148649

NULL

Week 32

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.91	2	0.45	6.20	0.003
Residuals	11.62	159	0.07
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 159) = 6.20, p = 0.003; Eta2 = 0.07, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.0991883	-0.0146736	0.2130501	0.1013685
8-1	0.2074630	0.0666367	0.3482892	0.0018296
8-2	0.1082747	-0.0256003	0.2421498	0.1382025

NULL

Week 33

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	0.70	2	0.35	6.55	0.002
Residuals	8.44	158	0.05
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 158) = 6.55, p = 0.002; Eta2 = 0.08, 95% CI [0.02, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1109694	0.0133113	0.2086274	0.0215593
8-1	0.1726119	0.0521780	0.2930458	0.0025206
8-2	0.0616425	-0.0530883	0.1763734	0.4135427

NULL

Week 34

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.97	2	0.98	14.61	< .001
Residuals	10.65	158	0.07
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 158) = 14.61, p < .001; Eta2 = 0.16, 95% CI [0.07, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1868757	0.0772095	0.2965418	0.0002523
8-1	0.2890028	0.1537603	0.4242454	0.0000035
8-2	0.1021271	-0.0267111	0.2309654	0.1491873

NULL

Week 35

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.89	2	0.95	12.28	< .001
Residuals	12.10	157	0.08
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and medium (F(2, 157) = 12.28, p < .001; Eta2 = 0.14, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1592373	0.0419567	0.2765179	0.0045259
8-1	0.2978010	0.1517626	0.4438395	0.0000098
8-2	0.1385637	-0.0006949	0.2778224	0.0514706

NULL

Week 36

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.29	2	1.14	16.11	< .001
Residuals	11.07	156	0.07
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(2, 156) = 16.11, p < .001; Eta2 = 0.17, 95% CI [0.09, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
2-1	0.1740027	0.0605118	0.2874936	0.0011201
8-1	0.3268445	0.1875465	0.4661424	0.0000004
8-2	0.1528418	0.0203129	0.2853706	0.0192950

NULL

Week 37

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.22	1	2.22	26.73	< .001
Residuals	6.98	84	0.08
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 84) = 26.73, p < .001; Eta2 = 0.24, 95% CI [0.12, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3303645	0.2032909	0.4574382	1.6e-06

NULL

Week 38

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.64	1	2.64	26.55	< .001
Residuals	8.25	83	0.10
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 83) = 26.55, p < .001; Eta2 = 0.24, 95% CI [0.12, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3614293	0.2219095	0.5009491	1.7e-06

NULL

Week 39

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.76	1	2.76	21.98	< .001
Residuals	10.17	81	0.13
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 81) = 21.98, p < .001; Eta2 = 0.21, 95% CI [0.10, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3746446	0.2156293	0.53366	1.1e-05

NULL

Week 40

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	2.53	1	2.53	21.96	< .001
Residuals	9.09	79	0.12
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 79) = 21.96, p < .001; Eta2 = 0.22, 95% CI [0.10, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3633933	0.2090524	0.5177342	1.14e-05

NULL

Week 41

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.90	1	1.90	16.80	< .001
Residuals	7.13	63	0.11
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 63) = 16.80, p < .001; Eta2 = 0.21, 95% CI [0.08, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3420326	0.1752971	0.5087681	0.0001208

NULL

Week 42

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.63	1	1.63	14.11	< .001
Residuals	6.94	60	0.12
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 60) = 14.11, p < .001; Eta2 = 0.19, 95% CI [0.06, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.324597	0.1517682	0.4974258	0.0003915

NULL

Week 43

Model Summary
Parameter	Sum_Squares	df	Mean_Square	F	p
datasetcode	1.66	1	1.66	16.60	< .001
Residuals	5.58	56	0.10
Anova Table (Type 1 tests)

Testing model assumption of residual normality with QQ plot

The ANOVA (formula: i ~ datasetcode) suggests that:

• The main effect of datasetcode is statistically significant and large (F(1, 56) = 16.60, p < .001; Eta2 = 0.23, 95% CI [0.09, 1.00])

Effect sizes were labelled following Field’s (2013) recommendations.
Statistical significance (P < 0.5) means we reject the null hypothesis that the means are equal.

Contrast	Difference	Lower CI	Upper CI	P-value
8-1	0.3386913	0.1721746	0.505208	0.0001468

NULL